Merge branch 'main' into refactor/query-params-nuqs

.claude/settings.json

@@ -4,6 +4,6 @@
     "context7@claude-plugins-official": true,
     "typescript-lsp@claude-plugins-official": true,
     "pyright-lsp@claude-plugins-official": true,
-    "ralph-wiggum@claude-plugins-official": true
+    "ralph-loop@claude-plugins-official": true
   }
 }

.claude/skills/skill-creator/SKILL.md

@@ -0,0 +1,355 @@
+---
+name: skill-creator
+description: Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.
+---
+
+# Skill Creator
+
+This skill provides guidance for creating effective skills.
+
+## About Skills
+
+Skills are modular, self-contained packages that extend Claude's capabilities by providing
+specialized knowledge, workflows, and tools. Think of them as "onboarding guides" for specific
+domains or tasks—they transform Claude from a general-purpose agent into a specialized agent
+equipped with procedural knowledge that no model can fully possess.
+
+### What Skills Provide
+
+1. Specialized workflows - Multi-step procedures for specific domains
+2. Tool integrations - Instructions for working with specific file formats or APIs
+3. Domain expertise - Company-specific knowledge, schemas, business logic
+4. Bundled resources - Scripts, references, and assets for complex and repetitive tasks
+
+## Core Principles
+
+### Concise is Key
+
+The context window is a public good. Skills share the context window with everything else Claude needs: system prompt, conversation history, other Skills' metadata, and the actual user request.
+
+**Default assumption: Claude is already very smart.** Only add context Claude doesn't already have. Challenge each piece of information: "Does Claude really need this explanation?" and "Does this paragraph justify its token cost?"
+
+Prefer concise examples over verbose explanations.
+
+### Set Appropriate Degrees of Freedom
+
+Match the level of specificity to the task's fragility and variability:
+
+**High freedom (text-based instructions)**: Use when multiple approaches are valid, decisions depend on context, or heuristics guide the approach.
+
+**Medium freedom (pseudocode or scripts with parameters)**: Use when a preferred pattern exists, some variation is acceptable, or configuration affects behavior.
+
+**Low freedom (specific scripts, few parameters)**: Use when operations are fragile and error-prone, consistency is critical, or a specific sequence must be followed.
+
+Think of Claude as exploring a path: a narrow bridge with cliffs needs specific guardrails (low freedom), while an open field allows many routes (high freedom).
+
+### Anatomy of a Skill
+
+Every skill consists of a required SKILL.md file and optional bundled resources:
+
+```
+skill-name/
+├── SKILL.md (required)
+│   ├── YAML frontmatter metadata (required)
+│   │   ├── name: (required)
+│   │   └── description: (required)
+│   └── Markdown instructions (required)
+└── Bundled Resources (optional)
+    ├── scripts/          - Executable code (Python/Bash/etc.)
+    ├── references/       - Documentation intended to be loaded into context as needed
+    └── assets/           - Files used in output (templates, icons, fonts, etc.)
+```
+
+#### SKILL.md (required)
+
+Every SKILL.md consists of:
+
+- **Frontmatter** (YAML): Contains `name` and `description` fields. These are the only fields that Claude reads to determine when the skill gets used, thus it is very important to be clear and comprehensive in describing what the skill is, and when it should be used.
+- **Body** (Markdown): Instructions and guidance for using the skill. Only loaded AFTER the skill triggers (if at all).
+
+#### Bundled Resources (optional)
+
+##### Scripts (`scripts/`)
+
+Executable code (Python/Bash/etc.) for tasks that require deterministic reliability or are repeatedly rewritten.
+
+- **When to include**: When the same code is being rewritten repeatedly or deterministic reliability is needed
+- **Example**: `scripts/rotate_pdf.py` for PDF rotation tasks
+- **Benefits**: Token efficient, deterministic, may be executed without loading into context
+- **Note**: Scripts may still need to be read by Claude for patching or environment-specific adjustments
+
+##### References (`references/`)
+
+Documentation and reference material intended to be loaded as needed into context to inform Claude's process and thinking.
+
+- **When to include**: For documentation that Claude should reference while working
+- **Examples**: `references/finance.md` for financial schemas, `references/mnda.md` for company NDA template, `references/policies.md` for company policies, `references/api_docs.md` for API specifications
+- **Use cases**: Database schemas, API documentation, domain knowledge, company policies, detailed workflow guides
+- **Benefits**: Keeps SKILL.md lean, loaded only when Claude determines it's needed
+- **Best practice**: If files are large (>10k words), include grep search patterns in SKILL.md
+- **Avoid duplication**: Information should live in either SKILL.md or references files, not both. Prefer references files for detailed information unless it's truly core to the skill—this keeps SKILL.md lean while making information discoverable without hogging the context window. Keep only essential procedural instructions and workflow guidance in SKILL.md; move detailed reference material, schemas, and examples to references files.
+
+##### Assets (`assets/`)
+
+Files not intended to be loaded into context, but rather used within the output Claude produces.
+
+- **When to include**: When the skill needs files that will be used in the final output
+- **Examples**: `assets/logo.png` for brand assets, `assets/slides.pptx` for PowerPoint templates, `assets/frontend-template/` for HTML/React boilerplate, `assets/font.ttf` for typography
+- **Use cases**: Templates, images, icons, boilerplate code, fonts, sample documents that get copied or modified
+- **Benefits**: Separates output resources from documentation, enables Claude to use files without loading them into context
+
+#### What to Not Include in a Skill
+
+A skill should only contain essential files that directly support its functionality. Do NOT create extraneous documentation or auxiliary files, including:
+
+- README.md
+- INSTALLATION_GUIDE.md
+- QUICK_REFERENCE.md
+- CHANGELOG.md
+- etc.
+
+The skill should only contain the information needed for an AI agent to do the job at hand. It should not contain auxilary context about the process that went into creating it, setup and testing procedures, user-facing documentation, etc. Creating additional documentation files just adds clutter and confusion.
+
+### Progressive Disclosure Design Principle
+
+Skills use a three-level loading system to manage context efficiently:
+
+1. **Metadata (name + description)** - Always in context (~100 words)
+2. **SKILL.md body** - When skill triggers (<5k words)
+3. **Bundled resources** - As needed by Claude (Unlimited because scripts can be executed without reading into context window)
+
+#### Progressive Disclosure Patterns
+
+Keep SKILL.md body to the essentials and under 500 lines to minimize context bloat. Split content into separate files when approaching this limit. When splitting out content into other files, it is very important to reference them from SKILL.md and describe clearly when to read them, to ensure the reader of the skill knows they exist and when to use them.
+
+**Key principle:** When a skill supports multiple variations, frameworks, or options, keep only the core workflow and selection guidance in SKILL.md. Move variant-specific details (patterns, examples, configuration) into separate reference files.
+
+**Pattern 1: High-level guide with references**
+
+```markdown
+# PDF Processing
+
+## Quick start
+
+Extract text with pdfplumber:
+[code example]
+
+## Advanced features
+
+- **Form filling**: See [FORMS.md](FORMS.md) for complete guide
+- **API reference**: See [REFERENCE.md](REFERENCE.md) for all methods
+- **Examples**: See [EXAMPLES.md](EXAMPLES.md) for common patterns
+```
+
+Claude loads FORMS.md, REFERENCE.md, or EXAMPLES.md only when needed.
+
+**Pattern 2: Domain-specific organization**
+
+For Skills with multiple domains, organize content by domain to avoid loading irrelevant context:
+
+```
+bigquery-skill/
+├── SKILL.md (overview and navigation)
+└── reference/
+    ├── finance.md (revenue, billing metrics)
+    ├── sales.md (opportunities, pipeline)
+    ├── product.md (API usage, features)
+    └── marketing.md (campaigns, attribution)
+```
+
+When a user asks about sales metrics, Claude only reads sales.md.
+
+Similarly, for skills supporting multiple frameworks or variants, organize by variant:
+
+```
+cloud-deploy/
+├── SKILL.md (workflow + provider selection)
+└── references/
+    ├── aws.md (AWS deployment patterns)
+    ├── gcp.md (GCP deployment patterns)
+    └── azure.md (Azure deployment patterns)
+```
+
+When the user chooses AWS, Claude only reads aws.md.
+
+**Pattern 3: Conditional details**
+
+Show basic content, link to advanced content:
+
+```markdown
+# DOCX Processing
+
+## Creating documents
+
+Use docx-js for new documents. See [DOCX-JS.md](DOCX-JS.md).
+
+## Editing documents
+
+For simple edits, modify the XML directly.
+
+**For tracked changes**: See [REDLINING.md](REDLINING.md)
+**For OOXML details**: See [OOXML.md](OOXML.md)
+```
+
+Claude reads REDLINING.md or OOXML.md only when the user needs those features.
+
+**Important guidelines:**
+
+- **Avoid deeply nested references** - Keep references one level deep from SKILL.md. All reference files should link directly from SKILL.md.
+- **Structure longer reference files** - For files longer than 100 lines, include a table of contents at the top so Claude can see the full scope when previewing.
+
+## Skill Creation Process
+
+Skill creation involves these steps:
+
+1. Understand the skill with concrete examples
+2. Plan reusable skill contents (scripts, references, assets)
+3. Initialize the skill (run init_skill.py)
+4. Edit the skill (implement resources and write SKILL.md)
+5. Package the skill (run package_skill.py)
+6. Iterate based on real usage
+
+Follow these steps in order, skipping only if there is a clear reason why they are not applicable.
+
+### Step 1: Understanding the Skill with Concrete Examples
+
+Skip this step only when the skill's usage patterns are already clearly understood. It remains valuable even when working with an existing skill.
+
+To create an effective skill, clearly understand concrete examples of how the skill will be used. This understanding can come from either direct user examples or generated examples that are validated with user feedback.
+
+For example, when building an image-editor skill, relevant questions include:
+
+- "What functionality should the image-editor skill support? Editing, rotating, anything else?"
+- "Can you give some examples of how this skill would be used?"
+- "I can imagine users asking for things like 'Remove the red-eye from this image' or 'Rotate this image'. Are there other ways you imagine this skill being used?"
+- "What would a user say that should trigger this skill?"
+
+To avoid overwhelming users, avoid asking too many questions in a single message. Start with the most important questions and follow up as needed for better effectiveness.
+
+Conclude this step when there is a clear sense of the functionality the skill should support.
+
+### Step 2: Planning the Reusable Skill Contents
+
+To turn concrete examples into an effective skill, analyze each example by:
+
+1. Considering how to execute on the example from scratch
+2. Identifying what scripts, references, and assets would be helpful when executing these workflows repeatedly
+
+Example: When building a `pdf-editor` skill to handle queries like "Help me rotate this PDF," the analysis shows:
+
+1. Rotating a PDF requires re-writing the same code each time
+2. A `scripts/rotate_pdf.py` script would be helpful to store in the skill
+
+Example: When designing a `frontend-webapp-builder` skill for queries like "Build me a todo app" or "Build me a dashboard to track my steps," the analysis shows:
+
+1. Writing a frontend webapp requires the same boilerplate HTML/React each time
+2. An `assets/hello-world/` template containing the boilerplate HTML/React project files would be helpful to store in the skill
+
+Example: When building a `big-query` skill to handle queries like "How many users have logged in today?" the analysis shows:
+
+1. Querying BigQuery requires re-discovering the table schemas and relationships each time
+2. A `references/schema.md` file documenting the table schemas would be helpful to store in the skill
+
+To establish the skill's contents, analyze each concrete example to create a list of the reusable resources to include: scripts, references, and assets.
+
+### Step 3: Initializing the Skill
+
+At this point, it is time to actually create the skill.
+
+Skip this step only if the skill being developed already exists, and iteration or packaging is needed. In this case, continue to the next step.
+
+When creating a new skill from scratch, always run the `init_skill.py` script. The script conveniently generates a new template skill directory that automatically includes everything a skill requires, making the skill creation process much more efficient and reliable.
+
+Usage:
+
+```bash
+scripts/init_skill.py <skill-name> --path <output-directory>
+```
+
+The script:
+
+- Creates the skill directory at the specified path
+- Generates a SKILL.md template with proper frontmatter and TODO placeholders
+- Creates example resource directories: `scripts/`, `references/`, and `assets/`
+- Adds example files in each directory that can be customized or deleted
+
+After initialization, customize or remove the generated SKILL.md and example files as needed.
+
+### Step 4: Edit the Skill
+
+When editing the (newly-generated or existing) skill, remember that the skill is being created for another instance of Claude to use. Include information that would be beneficial and non-obvious to Claude. Consider what procedural knowledge, domain-specific details, or reusable assets would help another Claude instance execute these tasks more effectively.
+
+#### Learn Proven Design Patterns
+
+Consult these helpful guides based on your skill's needs:
+
+- **Multi-step processes**: See references/workflows.md for sequential workflows and conditional logic
+- **Specific output formats or quality standards**: See references/output-patterns.md for template and example patterns
+
+These files contain established best practices for effective skill design.
+
+#### Start with Reusable Skill Contents
+
+To begin implementation, start with the reusable resources identified above: `scripts/`, `references/`, and `assets/` files. Note that this step may require user input. For example, when implementing a `brand-guidelines` skill, the user may need to provide brand assets or templates to store in `assets/`, or documentation to store in `references/`.
+
+Added scripts must be tested by actually running them to ensure there are no bugs and that the output matches what is expected. If there are many similar scripts, only a representative sample needs to be tested to ensure confidence that they all work while balancing time to completion.
+
+Any example files and directories not needed for the skill should be deleted. The initialization script creates example files in `scripts/`, `references/`, and `assets/` to demonstrate structure, but most skills won't need all of them.
+
+#### Update SKILL.md
+
+**Writing Guidelines:** Always use imperative/infinitive form.
+
+##### Frontmatter
+
+Write the YAML frontmatter with `name` and `description`:
+
+- `name`: The skill name
+- `description`: This is the primary triggering mechanism for your skill, and helps Claude understand when to use the skill.
+  - Include both what the Skill does and specific triggers/contexts for when to use it.
+  - Include all "when to use" information here - Not in the body. The body is only loaded after triggering, so "When to Use This Skill" sections in the body are not helpful to Claude.
+  - Example description for a `docx` skill: "Comprehensive document creation, editing, and analysis with support for tracked changes, comments, formatting preservation, and text extraction. Use when Claude needs to work with professional documents (.docx files) for: (1) Creating new documents, (2) Modifying or editing content, (3) Working with tracked changes, (4) Adding comments, or any other document tasks"
+
+Do not include any other fields in YAML frontmatter.
+
+##### Body
+
+Write instructions for using the skill and its bundled resources.
+
+### Step 5: Packaging a Skill
+
+Once development of the skill is complete, it must be packaged into a distributable .skill file that gets shared with the user. The packaging process automatically validates the skill first to ensure it meets all requirements:
+
+```bash
+scripts/package_skill.py <path/to/skill-folder>
+```
+
+Optional output directory specification:
+
+```bash
+scripts/package_skill.py <path/to/skill-folder> ./dist
+```
+
+The packaging script will:
+
+1. **Validate** the skill automatically, checking:
+
+   - YAML frontmatter format and required fields
+   - Skill naming conventions and directory structure
+   - Description completeness and quality
+   - File organization and resource references
+
+2. **Package** the skill if validation passes, creating a .skill file named after the skill (e.g., `my-skill.skill`) that includes all files and maintains the proper directory structure for distribution. The .skill file is a zip file with a .skill extension.
+
+If validation fails, the script will report the errors and exit without creating a package. Fix any validation errors and run the packaging command again.
+
+### Step 6: Iterate
+
+After testing the skill, users may request improvements. Often this happens right after using the skill, with fresh context of how the skill performed.
+
+**Iteration workflow:**
+
+1. Use the skill on real tasks
+2. Notice struggles or inefficiencies
+3. Identify how SKILL.md or bundled resources should be updated
+4. Implement changes and test again

.claude/skills/skill-creator/references/output-patterns.md

@@ -0,0 +1,86 @@
+# Output Patterns
+
+Use these patterns when skills need to produce consistent, high-quality output.
+
+## Template Pattern
+
+Provide templates for output format. Match the level of strictness to your needs.
+
+**For strict requirements (like API responses or data formats):**
+
+```markdown
+## Report structure
+
+ALWAYS use this exact template structure:
+
+# [Analysis Title]
+
+## Executive summary
+[One-paragraph overview of key findings]
+
+## Key findings
+- Finding 1 with supporting data
+- Finding 2 with supporting data
+- Finding 3 with supporting data
+
+## Recommendations
+1. Specific actionable recommendation
+2. Specific actionable recommendation
+```
+
+**For flexible guidance (when adaptation is useful):**
+
+```markdown
+## Report structure
+
+Here is a sensible default format, but use your best judgment:
+
+# [Analysis Title]
+
+## Executive summary
+[Overview]
+
+## Key findings
+[Adapt sections based on what you discover]
+
+## Recommendations
+[Tailor to the specific context]
+
+Adjust sections as needed for the specific analysis type.
+```
+
+## Examples Pattern
+
+For skills where output quality depends on seeing examples, provide input/output pairs:
+
+```markdown
+## Commit message format
+
+Generate commit messages following these examples:
+
+**Example 1:**
+Input: Added user authentication with JWT tokens
+Output:
+```
+
+feat(auth): implement JWT-based authentication
+
+Add login endpoint and token validation middleware
+
+```
+
+**Example 2:**
+Input: Fixed bug where dates displayed incorrectly in reports
+Output:
+```
+
+fix(reports): correct date formatting in timezone conversion
+
+Use UTC timestamps consistently across report generation
+
+```
+
+Follow this style: type(scope): brief description, then detailed explanation.
+```
+
+Examples help Claude understand the desired style and level of detail more clearly than descriptions alone.

.claude/skills/skill-creator/references/workflows.md

@@ -0,0 +1,28 @@
+# Workflow Patterns
+
+## Sequential Workflows
+
+For complex tasks, break operations into clear, sequential steps. It is often helpful to give Claude an overview of the process towards the beginning of SKILL.md:
+
+```markdown
+Filling a PDF form involves these steps:
+
+1. Analyze the form (run analyze_form.py)
+2. Create field mapping (edit fields.json)
+3. Validate mapping (run validate_fields.py)
+4. Fill the form (run fill_form.py)
+5. Verify output (run verify_output.py)
+```
+
+## Conditional Workflows
+
+For tasks with branching logic, guide Claude through decision points:
+
+```markdown
+1. Determine the modification type:
+   **Creating new content?** → Follow "Creation workflow" below
+   **Editing existing content?** → Follow "Editing workflow" below
+
+2. Creation workflow: [steps]
+3. Editing workflow: [steps]
+```

.claude/skills/skill-creator/scripts/init_skill.py

@@ -0,0 +1,300 @@
+#!/usr/bin/env python3
+"""
+Skill Initializer - Creates a new skill from template
+
+Usage:
+    init_skill.py <skill-name> --path <path>
+
+Examples:
+    init_skill.py my-new-skill --path skills/public
+    init_skill.py my-api-helper --path skills/private
+    init_skill.py custom-skill --path /custom/location
+"""
+
+import sys
+from pathlib import Path
+
+
+SKILL_TEMPLATE = """---
+name: {skill_name}
+description: [TODO: Complete and informative explanation of what the skill does and when to use it. Include WHEN to use this skill - specific scenarios, file types, or tasks that trigger it.]
+---
+
+# {skill_title}
+
+## Overview
+
+[TODO: 1-2 sentences explaining what this skill enables]
+
+## Structuring This Skill
+
+[TODO: Choose the structure that best fits this skill's purpose. Common patterns:
+
+**1. Workflow-Based** (best for sequential processes)
+- Works well when there are clear step-by-step procedures
+- Example: DOCX skill with "Workflow Decision Tree" → "Reading" → "Creating" → "Editing"
+- Structure: ## Overview → ## Workflow Decision Tree → ## Step 1 → ## Step 2...
+
+**2. Task-Based** (best for tool collections)
+- Works well when the skill offers different operations/capabilities
+- Example: PDF skill with "Quick Start" → "Merge PDFs" → "Split PDFs" → "Extract Text"
+- Structure: ## Overview → ## Quick Start → ## Task Category 1 → ## Task Category 2...
+
+**3. Reference/Guidelines** (best for standards or specifications)
+- Works well for brand guidelines, coding standards, or requirements
+- Example: Brand styling with "Brand Guidelines" → "Colors" → "Typography" → "Features"
+- Structure: ## Overview → ## Guidelines → ## Specifications → ## Usage...
+
+**4. Capabilities-Based** (best for integrated systems)
+- Works well when the skill provides multiple interrelated features
+- Example: Product Management with "Core Capabilities" → numbered capability list
+- Structure: ## Overview → ## Core Capabilities → ### 1. Feature → ### 2. Feature...
+
+Patterns can be mixed and matched as needed. Most skills combine patterns (e.g., start with task-based, add workflow for complex operations).
+
+Delete this entire "Structuring This Skill" section when done - it's just guidance.]
+
+## [TODO: Replace with the first main section based on chosen structure]
+
+[TODO: Add content here. See examples in existing skills:
+- Code samples for technical skills
+- Decision trees for complex workflows
+- Concrete examples with realistic user requests
+- References to scripts/templates/references as needed]
+
+## Resources
+
+This skill includes example resource directories that demonstrate how to organize different types of bundled resources:
+
+### scripts/
+Executable code (Python/Bash/etc.) that can be run directly to perform specific operations.
+
+**Examples from other skills:**
+- PDF skill: `fill_fillable_fields.py`, `extract_form_field_info.py` - utilities for PDF manipulation
+- DOCX skill: `document.py`, `utilities.py` - Python modules for document processing
+
+**Appropriate for:** Python scripts, shell scripts, or any executable code that performs automation, data processing, or specific operations.
+
+**Note:** Scripts may be executed without loading into context, but can still be read by Claude for patching or environment adjustments.
+
+### references/
+Documentation and reference material intended to be loaded into context to inform Claude's process and thinking.
+
+**Examples from other skills:**
+- Product management: `communication.md`, `context_building.md` - detailed workflow guides
+- BigQuery: API reference documentation and query examples
+- Finance: Schema documentation, company policies
+
+**Appropriate for:** In-depth documentation, API references, database schemas, comprehensive guides, or any detailed information that Claude should reference while working.
+
+### assets/
+Files not intended to be loaded into context, but rather used within the output Claude produces.
+
+**Examples from other skills:**
+- Brand styling: PowerPoint template files (.pptx), logo files
+- Frontend builder: HTML/React boilerplate project directories
+- Typography: Font files (.ttf, .woff2)
+
+**Appropriate for:** Templates, boilerplate code, document templates, images, icons, fonts, or any files meant to be copied or used in the final output.
+
+---
+
+**Any unneeded directories can be deleted.** Not every skill requires all three types of resources.
+"""
+
+EXAMPLE_SCRIPT = '''#!/usr/bin/env python3
+"""
+Example helper script for {skill_name}
+
+This is a placeholder script that can be executed directly.
+Replace with actual implementation or delete if not needed.
+
+Example real scripts from other skills:
+- pdf/scripts/fill_fillable_fields.py - Fills PDF form fields
+- pdf/scripts/convert_pdf_to_images.py - Converts PDF pages to images
+"""
+
+def main():
+    print("This is an example script for {skill_name}")
+    # TODO: Add actual script logic here
+    # This could be data processing, file conversion, API calls, etc.
+
+if __name__ == "__main__":
+    main()
+'''
+
+EXAMPLE_REFERENCE = """# Reference Documentation for {skill_title}
+
+This is a placeholder for detailed reference documentation.
+Replace with actual reference content or delete if not needed.
+
+Example real reference docs from other skills:
+- product-management/references/communication.md - Comprehensive guide for status updates
+- product-management/references/context_building.md - Deep-dive on gathering context
+- bigquery/references/ - API references and query examples
+
+## When Reference Docs Are Useful
+
+Reference docs are ideal for:
+- Comprehensive API documentation
+- Detailed workflow guides
+- Complex multi-step processes
+- Information too lengthy for main SKILL.md
+- Content that's only needed for specific use cases
+
+## Structure Suggestions
+
+### API Reference Example
+- Overview
+- Authentication
+- Endpoints with examples
+- Error codes
+- Rate limits
+
+### Workflow Guide Example
+- Prerequisites
+- Step-by-step instructions
+- Common patterns
+- Troubleshooting
+- Best practices
+"""
+
+EXAMPLE_ASSET = """# Example Asset File
+
+This placeholder represents where asset files would be stored.
+Replace with actual asset files (templates, images, fonts, etc.) or delete if not needed.
+
+Asset files are NOT intended to be loaded into context, but rather used within
+the output Claude produces.
+
+Example asset files from other skills:
+- Brand guidelines: logo.png, slides_template.pptx
+- Frontend builder: hello-world/ directory with HTML/React boilerplate
+- Typography: custom-font.ttf, font-family.woff2
+- Data: sample_data.csv, test_dataset.json
+
+## Common Asset Types
+
+- Templates: .pptx, .docx, boilerplate directories
+- Images: .png, .jpg, .svg, .gif
+- Fonts: .ttf, .otf, .woff, .woff2
+- Boilerplate code: Project directories, starter files
+- Icons: .ico, .svg
+- Data files: .csv, .json, .xml, .yaml
+
+Note: This is a text placeholder. Actual assets can be any file type.
+"""
+
+
+def title_case_skill_name(skill_name):
+    """Convert hyphenated skill name to Title Case for display."""
+    return " ".join(word.capitalize() for word in skill_name.split("-"))
+
+
+def init_skill(skill_name, path):
+    """
+    Initialize a new skill directory with template SKILL.md.
+
+    Args:
+        skill_name: Name of the skill
+        path: Path where the skill directory should be created
+
+    Returns:
+        Path to created skill directory, or None if error
+    """
+    # Determine skill directory path
+    skill_dir = Path(path).resolve() / skill_name
+
+    # Check if directory already exists
+    if skill_dir.exists():
+        print(f"❌ Error: Skill directory already exists: {skill_dir}")
+        return None
+
+    # Create skill directory
+    try:
+        skill_dir.mkdir(parents=True, exist_ok=False)
+        print(f"✅ Created skill directory: {skill_dir}")
+    except Exception as e:
+        print(f"❌ Error creating directory: {e}")
+        return None
+
+    # Create SKILL.md from template
+    skill_title = title_case_skill_name(skill_name)
+    skill_content = SKILL_TEMPLATE.format(skill_name=skill_name, skill_title=skill_title)
+
+    skill_md_path = skill_dir / "SKILL.md"
+    try:
+        skill_md_path.write_text(skill_content)
+        print("✅ Created SKILL.md")
+    except Exception as e:
+        print(f"❌ Error creating SKILL.md: {e}")
+        return None
+
+    # Create resource directories with example files
+    try:
+        # Create scripts/ directory with example script
+        scripts_dir = skill_dir / "scripts"
+        scripts_dir.mkdir(exist_ok=True)
+        example_script = scripts_dir / "example.py"
+        example_script.write_text(EXAMPLE_SCRIPT.format(skill_name=skill_name))
+        example_script.chmod(0o755)
+        print("✅ Created scripts/example.py")
+
+        # Create references/ directory with example reference doc
+        references_dir = skill_dir / "references"
+        references_dir.mkdir(exist_ok=True)
+        example_reference = references_dir / "api_reference.md"
+        example_reference.write_text(EXAMPLE_REFERENCE.format(skill_title=skill_title))
+        print("✅ Created references/api_reference.md")
+
+        # Create assets/ directory with example asset placeholder
+        assets_dir = skill_dir / "assets"
+        assets_dir.mkdir(exist_ok=True)
+        example_asset = assets_dir / "example_asset.txt"
+        example_asset.write_text(EXAMPLE_ASSET)
+        print("✅ Created assets/example_asset.txt")
+    except Exception as e:
+        print(f"❌ Error creating resource directories: {e}")
+        return None
+
+    # Print next steps
+    print(f"\n✅ Skill '{skill_name}' initialized successfully at {skill_dir}")
+    print("\nNext steps:")
+    print("1. Edit SKILL.md to complete the TODO items and update the description")
+    print("2. Customize or delete the example files in scripts/, references/, and assets/")
+    print("3. Run the validator when ready to check the skill structure")
+
+    return skill_dir
+
+
+def main():
+    if len(sys.argv) < 4 or sys.argv[2] != "--path":
+        print("Usage: init_skill.py <skill-name> --path <path>")
+        print("\nSkill name requirements:")
+        print("  - Hyphen-case identifier (e.g., 'data-analyzer')")
+        print("  - Lowercase letters, digits, and hyphens only")
+        print("  - Max 40 characters")
+        print("  - Must match directory name exactly")
+        print("\nExamples:")
+        print("  init_skill.py my-new-skill --path skills/public")
+        print("  init_skill.py my-api-helper --path skills/private")
+        print("  init_skill.py custom-skill --path /custom/location")
+        sys.exit(1)
+
+    skill_name = sys.argv[1]
+    path = sys.argv[3]
+
+    print(f"🚀 Initializing skill: {skill_name}")
+    print(f"   Location: {path}")
+    print()
+
+    result = init_skill(skill_name, path)
+
+    if result:
+        sys.exit(0)
+    else:
+        sys.exit(1)
+
+
+if __name__ == "__main__":
+    main()

.claude/skills/skill-creator/scripts/package_skill.py

@@ -0,0 +1,110 @@
+#!/usr/bin/env python3
+"""
+Skill Packager - Creates a distributable .skill file of a skill folder
+
+Usage:
+    python utils/package_skill.py <path/to/skill-folder> [output-directory]
+
+Example:
+    python utils/package_skill.py skills/public/my-skill
+    python utils/package_skill.py skills/public/my-skill ./dist
+"""
+
+import sys
+import zipfile
+from pathlib import Path
+from quick_validate import validate_skill
+
+
+def package_skill(skill_path, output_dir=None):
+    """
+    Package a skill folder into a .skill file.
+
+    Args:
+        skill_path: Path to the skill folder
+        output_dir: Optional output directory for the .skill file (defaults to current directory)
+
+    Returns:
+        Path to the created .skill file, or None if error
+    """
+    skill_path = Path(skill_path).resolve()
+
+    # Validate skill folder exists
+    if not skill_path.exists():
+        print(f"❌ Error: Skill folder not found: {skill_path}")
+        return None
+
+    if not skill_path.is_dir():
+        print(f"❌ Error: Path is not a directory: {skill_path}")
+        return None
+
+    # Validate SKILL.md exists
+    skill_md = skill_path / "SKILL.md"
+    if not skill_md.exists():
+        print(f"❌ Error: SKILL.md not found in {skill_path}")
+        return None
+
+    # Run validation before packaging
+    print("🔍 Validating skill...")
+    valid, message = validate_skill(skill_path)
+    if not valid:
+        print(f"❌ Validation failed: {message}")
+        print("   Please fix the validation errors before packaging.")
+        return None
+    print(f"✅ {message}\n")
+
+    # Determine output location
+    skill_name = skill_path.name
+    if output_dir:
+        output_path = Path(output_dir).resolve()
+        output_path.mkdir(parents=True, exist_ok=True)
+    else:
+        output_path = Path.cwd()
+
+    skill_filename = output_path / f"{skill_name}.skill"
+
+    # Create the .skill file (zip format)
+    try:
+        with zipfile.ZipFile(skill_filename, "w", zipfile.ZIP_DEFLATED) as zipf:
+            # Walk through the skill directory
+            for file_path in skill_path.rglob("*"):
+                if file_path.is_file():
+                    # Calculate the relative path within the zip
+                    arcname = file_path.relative_to(skill_path.parent)
+                    zipf.write(file_path, arcname)
+                    print(f"  Added: {arcname}")
+
+        print(f"\n✅ Successfully packaged skill to: {skill_filename}")
+        return skill_filename
+
+    except Exception as e:
+        print(f"❌ Error creating .skill file: {e}")
+        return None
+
+
+def main():
+    if len(sys.argv) < 2:
+        print("Usage: python utils/package_skill.py <path/to/skill-folder> [output-directory]")
+        print("\nExample:")
+        print("  python utils/package_skill.py skills/public/my-skill")
+        print("  python utils/package_skill.py skills/public/my-skill ./dist")
+        sys.exit(1)
+
+    skill_path = sys.argv[1]
+    output_dir = sys.argv[2] if len(sys.argv) > 2 else None
+
+    print(f"📦 Packaging skill: {skill_path}")
+    if output_dir:
+        print(f"   Output directory: {output_dir}")
+    print()
+
+    result = package_skill(skill_path, output_dir)
+
+    if result:
+        sys.exit(0)
+    else:
+        sys.exit(1)
+
+
+if __name__ == "__main__":
+    main()

.claude/skills/skill-creator/scripts/quick_validate.py

@@ -0,0 +1,97 @@
+#!/usr/bin/env python3
+"""
+Quick validation script for skills - minimal version
+"""
+
+import sys
+import os
+import re
+import yaml
+from pathlib import Path
+
+
+def validate_skill(skill_path):
+    """Basic validation of a skill"""
+    skill_path = Path(skill_path)
+
+    # Check SKILL.md exists
+    skill_md = skill_path / "SKILL.md"
+    if not skill_md.exists():
+        return False, "SKILL.md not found"
+
+    # Read and validate frontmatter
+    content = skill_md.read_text()
+    if not content.startswith("---"):
+        return False, "No YAML frontmatter found"
+
+    # Extract frontmatter
+    match = re.match(r"^---\n(.*?)\n---", content, re.DOTALL)
+    if not match:
+        return False, "Invalid frontmatter format"
+
+    frontmatter_text = match.group(1)
+
+    # Parse YAML frontmatter
+    try:
+        frontmatter = yaml.safe_load(frontmatter_text)
+        if not isinstance(frontmatter, dict):
+            return False, "Frontmatter must be a YAML dictionary"
+    except yaml.YAMLError as e:
+        return False, f"Invalid YAML in frontmatter: {e}"
+
+    # Define allowed properties
+    ALLOWED_PROPERTIES = {"name", "description", "license", "allowed-tools", "metadata"}
+
+    # Check for unexpected properties (excluding nested keys under metadata)
+    unexpected_keys = set(frontmatter.keys()) - ALLOWED_PROPERTIES
+    if unexpected_keys:
+        return False, (
+            f"Unexpected key(s) in SKILL.md frontmatter: {', '.join(sorted(unexpected_keys))}. "
+            f"Allowed properties are: {', '.join(sorted(ALLOWED_PROPERTIES))}"
+        )
+
+    # Check required fields
+    if "name" not in frontmatter:
+        return False, "Missing 'name' in frontmatter"
+    if "description" not in frontmatter:
+        return False, "Missing 'description' in frontmatter"
+
+    # Extract name for validation
+    name = frontmatter.get("name", "")
+    if not isinstance(name, str):
+        return False, f"Name must be a string, got {type(name).__name__}"
+    name = name.strip()
+    if name:
+        # Check naming convention (hyphen-case: lowercase with hyphens)
+        if not re.match(r"^[a-z0-9-]+$", name):
+            return False, f"Name '{name}' should be hyphen-case (lowercase letters, digits, and hyphens only)"
+        if name.startswith("-") or name.endswith("-") or "--" in name:
+            return False, f"Name '{name}' cannot start/end with hyphen or contain consecutive hyphens"
+        # Check name length (max 64 characters per spec)
+        if len(name) > 64:
+            return False, f"Name is too long ({len(name)} characters). Maximum is 64 characters."
+
+    # Extract and validate description
+    description = frontmatter.get("description", "")
+    if not isinstance(description, str):
+        return False, f"Description must be a string, got {type(description).__name__}"
+    description = description.strip()
+    if description:
+        # Check for angle brackets
+        if "<" in description or ">" in description:
+            return False, "Description cannot contain angle brackets (< or >)"
+        # Check description length (max 1024 characters per spec)
+        if len(description) > 1024:
+            return False, f"Description is too long ({len(description)} characters). Maximum is 1024 characters."
+
+    return True, "Skill is valid!"
+
+
+if __name__ == "__main__":
+    if len(sys.argv) != 2:
+        print("Usage: python quick_validate.py <skill_directory>")
+        sys.exit(1)
+
+    valid, message = validate_skill(sys.argv[1])
+    print(message)
+    sys.exit(0 if valid else 1)

api/.env.example

@@ -575,6 +575,10 @@ LOGSTORE_DUAL_WRITE_ENABLED=false
 # Enable dual-read fallback to SQL database when LogStore returns no results (default: true)
 # Useful for migration scenarios where historical data exists only in SQL database
 LOGSTORE_DUAL_READ_ENABLED=true
+# Control flag for whether to write the `graph` field to LogStore.
+# If LOGSTORE_ENABLE_PUT_GRAPH_FIELD is "true", write the full `graph` field;
+# otherwise write an empty {} instead. Defaults to writing the `graph` field.
+LOGSTORE_ENABLE_PUT_GRAPH_FIELD=true
 
 # Celery beat configuration
 CELERY_BEAT_SCHEDULER_TIME=1

api/.importlinter

@@ -53,7 +53,6 @@ ignore_imports =
     core.workflow.nodes.llm.llm_utils -> extensions.ext_database
     core.workflow.nodes.llm.node -> extensions.ext_database
     core.workflow.nodes.tool.tool_node -> extensions.ext_database
-    core.workflow.nodes.variable_assigner.common.impl -> extensions.ext_database
     core.workflow.graph_engine.command_channels.redis_channel -> extensions.ext_redis
     core.workflow.graph_engine.manager -> extensions.ext_redis
     core.workflow.nodes.knowledge_retrieval.knowledge_retrieval_node -> extensions.ext_redis

api/Dockerfile

@@ -50,16 +50,33 @@ WORKDIR /app/api
 
 # Create non-root user
 ARG dify_uid=1001
+ARG NODE_MAJOR=22
+ARG NODE_PACKAGE_VERSION=22.21.0-1nodesource1
+ARG NODESOURCE_KEY_FPR=6F71F525282841EEDAF851B42F59B5F99B1BE0B4
 RUN groupadd -r -g ${dify_uid} dify && \
     useradd -r -u ${dify_uid} -g ${dify_uid} -s /bin/bash dify && \
     chown -R dify:dify /app
 
 RUN \
     apt-get update \
+    && apt-get install -y --no-install-recommends \
+        ca-certificates \
+        curl \
+        gnupg \
+    && mkdir -p /etc/apt/keyrings \
+    && curl -fsSL https://deb.nodesource.com/gpgkey/nodesource-repo.gpg.key -o /tmp/nodesource.gpg \
+    && gpg --show-keys --with-colons /tmp/nodesource.gpg \
+        | awk -F: '/^fpr:/ {print $10}' \
+        | grep -Fx "${NODESOURCE_KEY_FPR}" \
+    && gpg --dearmor -o /etc/apt/keyrings/nodesource.gpg /tmp/nodesource.gpg \
+    && rm -f /tmp/nodesource.gpg \
+    && echo "deb [signed-by=/etc/apt/keyrings/nodesource.gpg] https://deb.nodesource.com/node_${NODE_MAJOR}.x nodistro main" \
+        > /etc/apt/sources.list.d/nodesource.list \
+    && apt-get update \
     # Install dependencies
     && apt-get install -y --no-install-recommends \
         # basic environment
-        curl nodejs \
+        nodejs=${NODE_PACKAGE_VERSION} \
         # for gmpy2 \
         libgmp-dev libmpfr-dev libmpc-dev \
         # For Security
@@ -79,7 +96,8 @@ COPY --from=packages --chown=dify:dify ${VIRTUAL_ENV} ${VIRTUAL_ENV}
 ENV PATH="${VIRTUAL_ENV}/bin:${PATH}"
 
 # Download nltk data
-RUN mkdir -p /usr/local/share/nltk_data && NLTK_DATA=/usr/local/share/nltk_data python -c "import nltk; nltk.download('punkt'); nltk.download('averaged_perceptron_tagger'); nltk.download('stopwords')" \
+RUN mkdir -p /usr/local/share/nltk_data \
+    && NLTK_DATA=/usr/local/share/nltk_data python -c "import nltk; from unstructured.nlp.tokenize import download_nltk_packages; nltk.download('punkt'); nltk.download('averaged_perceptron_tagger'); nltk.download('stopwords'); download_nltk_packages()" \
     && chmod -R 755 /usr/local/share/nltk_data
 
 ENV TIKTOKEN_CACHE_DIR=/app/api/.tiktoken_cache

api/commands.py

@@ -235,7 +235,7 @@ def migrate_annotation_vector_database():
                 if annotations:
                     for annotation in annotations:
                         document = Document(
-                            page_content=annotation.question,
+                            page_content=annotation.question_text,
                             metadata={"annotation_id": annotation.id, "app_id": app.id, "doc_id": annotation.id},
                         )
                         documents.append(document)

api/controllers/console/app/app.py

@@ -1,14 +1,16 @@
 import re
 import uuid
-from typing import Literal
+from datetime import datetime
+from typing import Any, Literal, TypeAlias
 
 from flask import request
-from flask_restx import Resource, fields, marshal, marshal_with
-from pydantic import BaseModel, Field, field_validator
+from flask_restx import Resource
+from pydantic import AliasChoices, BaseModel, ConfigDict, Field, computed_field, field_validator
 from sqlalchemy import select
 from sqlalchemy.orm import Session
 from werkzeug.exceptions import BadRequest
 
+from controllers.common.schema import register_schema_models
 from controllers.console import console_ns
 from controllers.console.app.wraps import get_app_model
 from controllers.console.wraps import (
@@ -19,27 +21,19 @@ from controllers.console.wraps import (
     is_admin_or_owner_required,
     setup_required,
 )
+from core.file import helpers as file_helpers
 from core.ops.ops_trace_manager import OpsTraceManager
 from core.workflow.enums import NodeType
 from extensions.ext_database import db
-from fields.app_fields import (
-    deleted_tool_fields,
-    model_config_fields,
-    model_config_partial_fields,
-    site_fields,
-    tag_fields,
-)
-from fields.workflow_fields import workflow_partial_fields as _workflow_partial_fields_dict
-from libs.helper import AppIconUrlField, TimestampField
 from libs.login import current_account_with_tenant, login_required
 from models import App, Workflow
+from models.model import IconType
 from services.app_dsl_service import AppDslService, ImportMode
 from services.app_service import AppService
 from services.enterprise.enterprise_service import EnterpriseService
 from services.feature_service import FeatureService
 
 ALLOW_CREATE_APP_MODES = ["chat", "agent-chat", "advanced-chat", "workflow", "completion"]
-DEFAULT_REF_TEMPLATE_SWAGGER_2_0 = "#/definitions/{model}"
 
 
 class AppListQuery(BaseModel):
@@ -192,124 +186,292 @@ class AppTracePayload(BaseModel):
         return value
 
 
-def reg(cls: type[BaseModel]):
-    console_ns.schema_model(cls.__name__, cls.model_json_schema(ref_template=DEFAULT_REF_TEMPLATE_SWAGGER_2_0))
+JSONValue: TypeAlias = Any
 
 
-reg(AppListQuery)
-reg(CreateAppPayload)
-reg(UpdateAppPayload)
-reg(CopyAppPayload)
-reg(AppExportQuery)
-reg(AppNamePayload)
-reg(AppIconPayload)
-reg(AppSiteStatusPayload)
-reg(AppApiStatusPayload)
-reg(AppTracePayload)
+class ResponseModel(BaseModel):
+    model_config = ConfigDict(
+        from_attributes=True,
+        extra="ignore",
+        populate_by_name=True,
+        serialize_by_alias=True,
+        protected_namespaces=(),
+    )
 
-# Register models for flask_restx to avoid dict type issues in Swagger
-# Register base models first
-tag_model = console_ns.model("Tag", tag_fields)
 
-workflow_partial_model = console_ns.model("WorkflowPartial", _workflow_partial_fields_dict)
+def _to_timestamp(value: datetime | int | None) -> int | None:
+    if isinstance(value, datetime):
+        return int(value.timestamp())
+    return value
 
-model_config_model = console_ns.model("ModelConfig", model_config_fields)
 
-model_config_partial_model = console_ns.model("ModelConfigPartial", model_config_partial_fields)
+def _build_icon_url(icon_type: str | IconType | None, icon: str | None) -> str | None:
+    if icon is None or icon_type is None:
+        return None
+    icon_type_value = icon_type.value if isinstance(icon_type, IconType) else str(icon_type)
+    if icon_type_value.lower() != IconType.IMAGE.value:
+        return None
+    return file_helpers.get_signed_file_url(icon)
 
-deleted_tool_model = console_ns.model("DeletedTool", deleted_tool_fields)
 
-site_model = console_ns.model("Site", site_fields)
+class Tag(ResponseModel):
+    id: str
+    name: str
+    type: str
 
-app_partial_model = console_ns.model(
-    "AppPartial",
-    {
-        "id": fields.String,
-        "name": fields.String,
-        "max_active_requests": fields.Raw(),
-        "description": fields.String(attribute="desc_or_prompt"),
-        "mode": fields.String(attribute="mode_compatible_with_agent"),
-        "icon_type": fields.String,
-        "icon": fields.String,
-        "icon_background": fields.String,
-        "icon_url": AppIconUrlField,
-        "model_config": fields.Nested(model_config_partial_model, attribute="app_model_config", allow_null=True),
-        "workflow": fields.Nested(workflow_partial_model, allow_null=True),
-        "use_icon_as_answer_icon": fields.Boolean,
-        "created_by": fields.String,
-        "created_at": TimestampField,
-        "updated_by": fields.String,
-        "updated_at": TimestampField,
-        "tags": fields.List(fields.Nested(tag_model)),
-        "access_mode": fields.String,
-        "create_user_name": fields.String,
-        "author_name": fields.String,
-        "has_draft_trigger": fields.Boolean,
-    },
-)
 
-app_detail_model = console_ns.model(
-    "AppDetail",
-    {
-        "id": fields.String,
-        "name": fields.String,
-        "description": fields.String,
-        "mode": fields.String(attribute="mode_compatible_with_agent"),
-        "icon": fields.String,
-        "icon_background": fields.String,
-        "enable_site": fields.Boolean,
-        "enable_api": fields.Boolean,
-        "model_config": fields.Nested(model_config_model, attribute="app_model_config", allow_null=True),
-        "workflow": fields.Nested(workflow_partial_model, allow_null=True),
-        "tracing": fields.Raw,
-        "use_icon_as_answer_icon": fields.Boolean,
-        "created_by": fields.String,
-        "created_at": TimestampField,
-        "updated_by": fields.String,
-        "updated_at": TimestampField,
-        "access_mode": fields.String,
-        "tags": fields.List(fields.Nested(tag_model)),
-    },
-)
+class WorkflowPartial(ResponseModel):
+    id: str
+    created_by: str | None = None
+    created_at: int | None = None
+    updated_by: str | None = None
+    updated_at: int | None = None
 
-app_detail_with_site_model = console_ns.model(
-    "AppDetailWithSite",
-    {
-        "id": fields.String,
-        "name": fields.String,
-        "description": fields.String,
-        "mode": fields.String(attribute="mode_compatible_with_agent"),
-        "icon_type": fields.String,
-        "icon": fields.String,
-        "icon_background": fields.String,
-        "icon_url": AppIconUrlField,
-        "enable_site": fields.Boolean,
-        "enable_api": fields.Boolean,
-        "model_config": fields.Nested(model_config_model, attribute="app_model_config", allow_null=True),
-        "workflow": fields.Nested(workflow_partial_model, allow_null=True),
-        "api_base_url": fields.String,
-        "use_icon_as_answer_icon": fields.Boolean,
-        "max_active_requests": fields.Integer,
-        "created_by": fields.String,
-        "created_at": TimestampField,
-        "updated_by": fields.String,
-        "updated_at": TimestampField,
-        "deleted_tools": fields.List(fields.Nested(deleted_tool_model)),
-        "access_mode": fields.String,
-        "tags": fields.List(fields.Nested(tag_model)),
-        "site": fields.Nested(site_model),
-    },
-)
+    @field_validator("created_at", "updated_at", mode="before")
+    @classmethod
+    def _normalize_timestamp(cls, value: datetime | int | None) -> int | None:
+        return _to_timestamp(value)
 
-app_pagination_model = console_ns.model(
-    "AppPagination",
-    {
-        "page": fields.Integer,
-        "limit": fields.Integer(attribute="per_page"),
-        "total": fields.Integer,
-        "has_more": fields.Boolean(attribute="has_next"),
-        "data": fields.List(fields.Nested(app_partial_model), attribute="items"),
-    },
+
+class ModelConfigPartial(ResponseModel):
+    model: JSONValue | None = Field(default=None, validation_alias=AliasChoices("model_dict", "model"))
+    pre_prompt: str | None = None
+    created_by: str | None = None
+    created_at: int | None = None
+    updated_by: str | None = None
+    updated_at: int | None = None
+
+    @field_validator("created_at", "updated_at", mode="before")
+    @classmethod
+    def _normalize_timestamp(cls, value: datetime | int | None) -> int | None:
+        return _to_timestamp(value)
+
+
+class ModelConfig(ResponseModel):
+    opening_statement: str | None = None
+    suggested_questions: JSONValue | None = Field(
+        default=None, validation_alias=AliasChoices("suggested_questions_list", "suggested_questions")
+    )
+    suggested_questions_after_answer: JSONValue | None = Field(
+        default=None,
+        validation_alias=AliasChoices("suggested_questions_after_answer_dict", "suggested_questions_after_answer"),
+    )
+    speech_to_text: JSONValue | None = Field(
+        default=None, validation_alias=AliasChoices("speech_to_text_dict", "speech_to_text")
+    )
+    text_to_speech: JSONValue | None = Field(
+        default=None, validation_alias=AliasChoices("text_to_speech_dict", "text_to_speech")
+    )
+    retriever_resource: JSONValue | None = Field(
+        default=None, validation_alias=AliasChoices("retriever_resource_dict", "retriever_resource")
+    )
+    annotation_reply: JSONValue | None = Field(
+        default=None, validation_alias=AliasChoices("annotation_reply_dict", "annotation_reply")
+    )
+    more_like_this: JSONValue | None = Field(
+        default=None, validation_alias=AliasChoices("more_like_this_dict", "more_like_this")
+    )
+    sensitive_word_avoidance: JSONValue | None = Field(
+        default=None, validation_alias=AliasChoices("sensitive_word_avoidance_dict", "sensitive_word_avoidance")
+    )
+    external_data_tools: JSONValue | None = Field(
+        default=None, validation_alias=AliasChoices("external_data_tools_list", "external_data_tools")
+    )
+    model: JSONValue | None = Field(default=None, validation_alias=AliasChoices("model_dict", "model"))
+    user_input_form: JSONValue | None = Field(
+        default=None, validation_alias=AliasChoices("user_input_form_list", "user_input_form")
+    )
+    dataset_query_variable: str | None = None
+    pre_prompt: str | None = None
+    agent_mode: JSONValue | None = Field(default=None, validation_alias=AliasChoices("agent_mode_dict", "agent_mode"))
+    prompt_type: str | None = None
+    chat_prompt_config: JSONValue | None = Field(
+        default=None, validation_alias=AliasChoices("chat_prompt_config_dict", "chat_prompt_config")
+    )
+    completion_prompt_config: JSONValue | None = Field(
+        default=None, validation_alias=AliasChoices("completion_prompt_config_dict", "completion_prompt_config")
+    )
+    dataset_configs: JSONValue | None = Field(
+        default=None, validation_alias=AliasChoices("dataset_configs_dict", "dataset_configs")
+    )
+    file_upload: JSONValue | None = Field(
+        default=None, validation_alias=AliasChoices("file_upload_dict", "file_upload")
+    )
+    created_by: str | None = None
+    created_at: int | None = None
+    updated_by: str | None = None
+    updated_at: int | None = None
+
+    @field_validator("created_at", "updated_at", mode="before")
+    @classmethod
+    def _normalize_timestamp(cls, value: datetime | int | None) -> int | None:
+        return _to_timestamp(value)
+
+
+class Site(ResponseModel):
+    access_token: str | None = Field(default=None, validation_alias="code")
+    code: str | None = None
+    title: str | None = None
+    icon_type: str | IconType | None = None
+    icon: str | None = None
+    icon_background: str | None = None
+    description: str | None = None
+    default_language: str | None = None
+    chat_color_theme: str | None = None
+    chat_color_theme_inverted: bool | None = None
+    customize_domain: str | None = None
+    copyright: str | None = None
+    privacy_policy: str | None = None
+    custom_disclaimer: str | None = None
+    customize_token_strategy: str | None = None
+    prompt_public: bool | None = None
+    app_base_url: str | None = None
+    show_workflow_steps: bool | None = None
+    use_icon_as_answer_icon: bool | None = None
+    created_by: str | None = None
+    created_at: int | None = None
+    updated_by: str | None = None
+    updated_at: int | None = None
+
+    @computed_field(return_type=str | None)  # type: ignore
+    @property
+    def icon_url(self) -> str | None:
+        return _build_icon_url(self.icon_type, self.icon)
+
+    @field_validator("icon_type", mode="before")
+    @classmethod
+    def _normalize_icon_type(cls, value: str | IconType | None) -> str | None:
+        if isinstance(value, IconType):
+            return value.value
+        return value
+
+    @field_validator("created_at", "updated_at", mode="before")
+    @classmethod
+    def _normalize_timestamp(cls, value: datetime | int | None) -> int | None:
+        return _to_timestamp(value)
+
+
+class DeletedTool(ResponseModel):
+    type: str
+    tool_name: str
+    provider_id: str
+
+
+class AppPartial(ResponseModel):
+    id: str
+    name: str
+    max_active_requests: int | None = None
+    description: str | None = Field(default=None, validation_alias=AliasChoices("desc_or_prompt", "description"))
+    mode: str = Field(validation_alias="mode_compatible_with_agent")
+    icon_type: str | None = None
+    icon: str | None = None
+    icon_background: str | None = None
+    model_config_: ModelConfigPartial | None = Field(
+        default=None,
+        validation_alias=AliasChoices("app_model_config", "model_config"),
+        alias="model_config",
+    )
+    workflow: WorkflowPartial | None = None
+    use_icon_as_answer_icon: bool | None = None
+    created_by: str | None = None
+    created_at: int | None = None
+    updated_by: str | None = None
+    updated_at: int | None = None
+    tags: list[Tag] = Field(default_factory=list)
+    access_mode: str | None = None
+    create_user_name: str | None = None
+    author_name: str | None = None
+    has_draft_trigger: bool | None = None
+
+    @computed_field(return_type=str | None)  # type: ignore
+    @property
+    def icon_url(self) -> str | None:
+        return _build_icon_url(self.icon_type, self.icon)
+
+    @field_validator("created_at", "updated_at", mode="before")
+    @classmethod
+    def _normalize_timestamp(cls, value: datetime | int | None) -> int | None:
+        return _to_timestamp(value)
+
+
+class AppDetail(ResponseModel):
+    id: str
+    name: str
+    description: str | None = None
+    mode: str = Field(validation_alias="mode_compatible_with_agent")
+    icon: str | None = None
+    icon_background: str | None = None
+    enable_site: bool
+    enable_api: bool
+    model_config_: ModelConfig | None = Field(
+        default=None,
+        validation_alias=AliasChoices("app_model_config", "model_config"),
+        alias="model_config",
+    )
+    workflow: WorkflowPartial | None = None
+    tracing: JSONValue | None = None
+    use_icon_as_answer_icon: bool | None = None
+    created_by: str | None = None
+    created_at: int | None = None
+    updated_by: str | None = None
+    updated_at: int | None = None
+    access_mode: str | None = None
+    tags: list[Tag] = Field(default_factory=list)
+
+    @field_validator("created_at", "updated_at", mode="before")
+    @classmethod
+    def _normalize_timestamp(cls, value: datetime | int | None) -> int | None:
+        return _to_timestamp(value)
+
+
+class AppDetailWithSite(AppDetail):
+    icon_type: str | None = None
+    api_base_url: str | None = None
+    max_active_requests: int | None = None
+    deleted_tools: list[DeletedTool] = Field(default_factory=list)
+    site: Site | None = None
+
+    @computed_field(return_type=str | None)  # type: ignore
+    @property
+    def icon_url(self) -> str | None:
+        return _build_icon_url(self.icon_type, self.icon)
+
+
+class AppPagination(ResponseModel):
+    page: int
+    limit: int = Field(validation_alias=AliasChoices("per_page", "limit"))
+    total: int
+    has_more: bool = Field(validation_alias=AliasChoices("has_next", "has_more"))
+    data: list[AppPartial] = Field(validation_alias=AliasChoices("items", "data"))
+
+
+class AppExportResponse(ResponseModel):
+    data: str
+
+
+register_schema_models(
+    console_ns,
+    AppListQuery,
+    CreateAppPayload,
+    UpdateAppPayload,
+    CopyAppPayload,
+    AppExportQuery,
+    AppNamePayload,
+    AppIconPayload,
+    AppSiteStatusPayload,
+    AppApiStatusPayload,
+    AppTracePayload,
+    Tag,
+    WorkflowPartial,
+    ModelConfigPartial,
+    ModelConfig,
+    Site,
+    DeletedTool,
+    AppPartial,
+    AppDetail,
+    AppDetailWithSite,
+    AppPagination,
+    AppExportResponse,
 )
 
 
@@ -318,7 +480,7 @@ class AppListApi(Resource):
     @console_ns.doc("list_apps")
     @console_ns.doc(description="Get list of applications with pagination and filtering")
     @console_ns.expect(console_ns.models[AppListQuery.__name__])
-    @console_ns.response(200, "Success", app_pagination_model)
+    @console_ns.response(200, "Success", console_ns.models[AppPagination.__name__])
     @setup_required
     @login_required
     @account_initialization_required
@@ -334,7 +496,8 @@ class AppListApi(Resource):
         app_service = AppService()
         app_pagination = app_service.get_paginate_apps(current_user.id, current_tenant_id, args_dict)
         if not app_pagination:
-            return {"data": [], "total": 0, "page": 1, "limit": 20, "has_more": False}
+            empty = AppPagination(page=args.page, limit=args.limit, total=0, has_more=False, data=[])
+            return empty.model_dump(mode="json"), 200
 
         if FeatureService.get_system_features().webapp_auth.enabled:
             app_ids = [str(app.id) for app in app_pagination.items]
@@ -378,18 +541,18 @@ class AppListApi(Resource):
         for app in app_pagination.items:
             app.has_draft_trigger = str(app.id) in draft_trigger_app_ids
 
-        return marshal(app_pagination, app_pagination_model), 200
+        pagination_model = AppPagination.model_validate(app_pagination, from_attributes=True)
+        return pagination_model.model_dump(mode="json"), 200
 
     @console_ns.doc("create_app")
     @console_ns.doc(description="Create a new application")
     @console_ns.expect(console_ns.models[CreateAppPayload.__name__])
-    @console_ns.response(201, "App created successfully", app_detail_model)
+    @console_ns.response(201, "App created successfully", console_ns.models[AppDetail.__name__])
     @console_ns.response(403, "Insufficient permissions")
     @console_ns.response(400, "Invalid request parameters")
     @setup_required
     @login_required
     @account_initialization_required
-    @marshal_with(app_detail_model)
     @cloud_edition_billing_resource_check("apps")
     @edit_permission_required
     def post(self):
@@ -399,8 +562,8 @@ class AppListApi(Resource):
 
         app_service = AppService()
         app = app_service.create_app(current_tenant_id, args.model_dump(), current_user)
-
-        return app, 201
+        app_detail = AppDetail.model_validate(app, from_attributes=True)
+        return app_detail.model_dump(mode="json"), 201
 
 
 @console_ns.route("/apps/<uuid:app_id>")
@@ -408,13 +571,12 @@ class AppApi(Resource):
     @console_ns.doc("get_app_detail")
     @console_ns.doc(description="Get application details")
     @console_ns.doc(params={"app_id": "Application ID"})
-    @console_ns.response(200, "Success", app_detail_with_site_model)
+    @console_ns.response(200, "Success", console_ns.models[AppDetailWithSite.__name__])
     @setup_required
     @login_required
     @account_initialization_required
     @enterprise_license_required
-    @get_app_model
-    @marshal_with(app_detail_with_site_model)
+    @get_app_model(mode=None)
     def get(self, app_model):
         """Get app detail"""
         app_service = AppService()
@@ -425,21 +587,21 @@ class AppApi(Resource):
             app_setting = EnterpriseService.WebAppAuth.get_app_access_mode_by_id(app_id=str(app_model.id))
             app_model.access_mode = app_setting.access_mode
 
-        return app_model
+        response_model = AppDetailWithSite.model_validate(app_model, from_attributes=True)
+        return response_model.model_dump(mode="json")
 
     @console_ns.doc("update_app")
     @console_ns.doc(description="Update application details")
     @console_ns.doc(params={"app_id": "Application ID"})
     @console_ns.expect(console_ns.models[UpdateAppPayload.__name__])
-    @console_ns.response(200, "App updated successfully", app_detail_with_site_model)
+    @console_ns.response(200, "App updated successfully", console_ns.models[AppDetailWithSite.__name__])
     @console_ns.response(403, "Insufficient permissions")
     @console_ns.response(400, "Invalid request parameters")
     @setup_required
     @login_required
     @account_initialization_required
-    @get_app_model
+    @get_app_model(mode=None)
     @edit_permission_required
-    @marshal_with(app_detail_with_site_model)
     def put(self, app_model):
         """Update app"""
         args = UpdateAppPayload.model_validate(console_ns.payload)
@@ -456,8 +618,8 @@ class AppApi(Resource):
             "max_active_requests": args.max_active_requests or 0,
         }
         app_model = app_service.update_app(app_model, args_dict)
-
-        return app_model
+        response_model = AppDetailWithSite.model_validate(app_model, from_attributes=True)
+        return response_model.model_dump(mode="json")
 
     @console_ns.doc("delete_app")
     @console_ns.doc(description="Delete application")
@@ -483,14 +645,13 @@ class AppCopyApi(Resource):
     @console_ns.doc(description="Create a copy of an existing application")
     @console_ns.doc(params={"app_id": "Application ID to copy"})
     @console_ns.expect(console_ns.models[CopyAppPayload.__name__])
-    @console_ns.response(201, "App copied successfully", app_detail_with_site_model)
+    @console_ns.response(201, "App copied successfully", console_ns.models[AppDetailWithSite.__name__])
     @console_ns.response(403, "Insufficient permissions")
     @setup_required
     @login_required
     @account_initialization_required
-    @get_app_model
+    @get_app_model(mode=None)
     @edit_permission_required
-    @marshal_with(app_detail_with_site_model)
     def post(self, app_model):
         """Copy app"""
         # The role of the current user in the ta table must be admin, owner, or editor
@@ -516,7 +677,8 @@ class AppCopyApi(Resource):
             stmt = select(App).where(App.id == result.app_id)
             app = session.scalar(stmt)
 
-        return app, 201
+        response_model = AppDetailWithSite.model_validate(app, from_attributes=True)
+        return response_model.model_dump(mode="json"), 201
 
 
 @console_ns.route("/apps/<uuid:app_id>/export")
@@ -525,11 +687,7 @@ class AppExportApi(Resource):
     @console_ns.doc(description="Export application configuration as DSL")
     @console_ns.doc(params={"app_id": "Application ID to export"})
     @console_ns.expect(console_ns.models[AppExportQuery.__name__])
-    @console_ns.response(
-        200,
-        "App exported successfully",
-        console_ns.model("AppExportResponse", {"data": fields.String(description="DSL export data")}),
-    )
+    @console_ns.response(200, "App exported successfully", console_ns.models[AppExportResponse.__name__])
     @console_ns.response(403, "Insufficient permissions")
     @get_app_model
     @setup_required
@@ -540,13 +698,14 @@ class AppExportApi(Resource):
         """Export app"""
         args = AppExportQuery.model_validate(request.args.to_dict(flat=True))  # type: ignore
 
-        return {
-            "data": AppDslService.export_dsl(
+        payload = AppExportResponse(
+            data=AppDslService.export_dsl(
                 app_model=app_model,
                 include_secret=args.include_secret,
                 workflow_id=args.workflow_id,
             )
-        }
+        )
+        return payload.model_dump(mode="json")
 
 
 @console_ns.route("/apps/<uuid:app_id>/name")
@@ -555,20 +714,19 @@ class AppNameApi(Resource):
     @console_ns.doc(description="Check if app name is available")
     @console_ns.doc(params={"app_id": "Application ID"})
     @console_ns.expect(console_ns.models[AppNamePayload.__name__])
-    @console_ns.response(200, "Name availability checked")
+    @console_ns.response(200, "Name availability checked", console_ns.models[AppDetail.__name__])
     @setup_required
     @login_required
     @account_initialization_required
-    @get_app_model
-    @marshal_with(app_detail_model)
+    @get_app_model(mode=None)
     @edit_permission_required
     def post(self, app_model):
         args = AppNamePayload.model_validate(console_ns.payload)
 
         app_service = AppService()
         app_model = app_service.update_app_name(app_model, args.name)
-
-        return app_model
+        response_model = AppDetail.model_validate(app_model, from_attributes=True)
+        return response_model.model_dump(mode="json")
 
 
 @console_ns.route("/apps/<uuid:app_id>/icon")
@@ -582,16 +740,15 @@ class AppIconApi(Resource):
     @setup_required
     @login_required
     @account_initialization_required
-    @get_app_model
-    @marshal_with(app_detail_model)
+    @get_app_model(mode=None)
     @edit_permission_required
     def post(self, app_model):
         args = AppIconPayload.model_validate(console_ns.payload or {})
 
         app_service = AppService()
         app_model = app_service.update_app_icon(app_model, args.icon or "", args.icon_background or "")
-
-        return app_model
+        response_model = AppDetail.model_validate(app_model, from_attributes=True)
+        return response_model.model_dump(mode="json")
 
 
 @console_ns.route("/apps/<uuid:app_id>/site-enable")
@@ -600,21 +757,20 @@ class AppSiteStatus(Resource):
     @console_ns.doc(description="Enable or disable app site")
     @console_ns.doc(params={"app_id": "Application ID"})
     @console_ns.expect(console_ns.models[AppSiteStatusPayload.__name__])
-    @console_ns.response(200, "Site status updated successfully", app_detail_model)
+    @console_ns.response(200, "Site status updated successfully", console_ns.models[AppDetail.__name__])
     @console_ns.response(403, "Insufficient permissions")
     @setup_required
     @login_required
     @account_initialization_required
-    @get_app_model
-    @marshal_with(app_detail_model)
+    @get_app_model(mode=None)
     @edit_permission_required
     def post(self, app_model):
         args = AppSiteStatusPayload.model_validate(console_ns.payload)
 
         app_service = AppService()
         app_model = app_service.update_app_site_status(app_model, args.enable_site)
-
-        return app_model
+        response_model = AppDetail.model_validate(app_model, from_attributes=True)
+        return response_model.model_dump(mode="json")
 
 
 @console_ns.route("/apps/<uuid:app_id>/api-enable")
@@ -623,21 +779,20 @@ class AppApiStatus(Resource):
     @console_ns.doc(description="Enable or disable app API")
     @console_ns.doc(params={"app_id": "Application ID"})
     @console_ns.expect(console_ns.models[AppApiStatusPayload.__name__])
-    @console_ns.response(200, "API status updated successfully", app_detail_model)
+    @console_ns.response(200, "API status updated successfully", console_ns.models[AppDetail.__name__])
     @console_ns.response(403, "Insufficient permissions")
     @setup_required
     @login_required
     @is_admin_or_owner_required
     @account_initialization_required
-    @get_app_model
-    @marshal_with(app_detail_model)
+    @get_app_model(mode=None)
     def post(self, app_model):
         args = AppApiStatusPayload.model_validate(console_ns.payload)
 
         app_service = AppService()
         app_model = app_service.update_app_api_status(app_model, args.enable_api)
-
-        return app_model
+        response_model = AppDetail.model_validate(app_model, from_attributes=True)
+        return response_model.model_dump(mode="json")
 
 
 @console_ns.route("/apps/<uuid:app_id>/trace")

api/controllers/console/app/conversation.py

@@ -348,10 +348,13 @@ class CompletionConversationApi(Resource):
         )
 
         if args.keyword:
+            from libs.helper import escape_like_pattern
+
+            escaped_keyword = escape_like_pattern(args.keyword)
             query = query.join(Message, Message.conversation_id == Conversation.id).where(
                 or_(
-                    Message.query.ilike(f"%{args.keyword}%"),
-                    Message.answer.ilike(f"%{args.keyword}%"),
+                    Message.query.ilike(f"%{escaped_keyword}%", escape="\\"),
+                    Message.answer.ilike(f"%{escaped_keyword}%", escape="\\"),
                 )
             )
 
@@ -460,7 +463,10 @@ class ChatConversationApi(Resource):
         query = sa.select(Conversation).where(Conversation.app_id == app_model.id, Conversation.is_deleted.is_(False))
 
         if args.keyword:
-            keyword_filter = f"%{args.keyword}%"
+            from libs.helper import escape_like_pattern
+
+            escaped_keyword = escape_like_pattern(args.keyword)
+            keyword_filter = f"%{escaped_keyword}%"
             query = (
                 query.join(
                     Message,
@@ -469,11 +475,11 @@ class ChatConversationApi(Resource):
                 .join(subquery, subquery.c.conversation_id == Conversation.id)
                 .where(
                     or_(
-                        Message.query.ilike(keyword_filter),
-                        Message.answer.ilike(keyword_filter),
-                        Conversation.name.ilike(keyword_filter),
-                        Conversation.introduction.ilike(keyword_filter),
-                        subquery.c.from_end_user_session_id.ilike(keyword_filter),
+                        Message.query.ilike(keyword_filter, escape="\\"),
+                        Message.answer.ilike(keyword_filter, escape="\\"),
+                        Conversation.name.ilike(keyword_filter, escape="\\"),
+                        Conversation.introduction.ilike(keyword_filter, escape="\\"),
+                        subquery.c.from_end_user_session_id.ilike(keyword_filter, escape="\\"),
                     ),
                 )
                 .group_by(Conversation.id)

api/controllers/console/auth/login.py

@@ -1,3 +1,5 @@
+from typing import Any
+
 import flask_login
 from flask import make_response, request
 from flask_restx import Resource
@@ -96,14 +98,13 @@ class LoginApi(Resource):
         if is_login_error_rate_limit:
             raise EmailPasswordLoginLimitError()
 
-        # TODO: why invitation is re-assigned with different type?
-        invitation = args.invite_token  # type: ignore
-        if invitation:
-            invitation = RegisterService.get_invitation_if_token_valid(None, args.email, invitation)  # type: ignore
+        invitation_data: dict[str, Any] | None = None
+        if args.invite_token:
+            invitation_data = RegisterService.get_invitation_if_token_valid(None, args.email, args.invite_token)
 
         try:
-            if invitation:
-                data = invitation.get("data", {})  # type: ignore
+            if invitation_data:
+                data = invitation_data.get("data", {})
                 invitee_email = data.get("email") if data else None
                 if invitee_email != args.email:
                     raise InvalidEmailError()

api/controllers/console/datasets/datasets_segments.py

@@ -30,6 +30,7 @@ from core.model_runtime.entities.model_entities import ModelType
 from extensions.ext_database import db
 from extensions.ext_redis import redis_client
 from fields.segment_fields import child_chunk_fields, segment_fields
+from libs.helper import escape_like_pattern
 from libs.login import current_account_with_tenant, login_required
 from models.dataset import ChildChunk, DocumentSegment
 from models.model import UploadFile
@@ -145,6 +146,8 @@ class DatasetDocumentSegmentListApi(Resource):
             query = query.where(DocumentSegment.hit_count >= hit_count_gte)
 
         if keyword:
+            # Escape special characters in keyword to prevent SQL injection via LIKE wildcards
+            escaped_keyword = escape_like_pattern(keyword)
             # Search in both content and keywords fields
             # Use database-specific methods for JSON array search
             if dify_config.SQLALCHEMY_DATABASE_URI_SCHEME == "postgresql":
@@ -156,15 +159,15 @@ class DatasetDocumentSegmentListApi(Resource):
                         .scalar_subquery()
                     ),
                     ",",
-                ).ilike(f"%{keyword}%")
+                ).ilike(f"%{escaped_keyword}%", escape="\\")
             else:
                 # MySQL: Cast JSON to string for pattern matching
                 # MySQL stores Chinese text directly in JSON without Unicode escaping
-                keywords_condition = cast(DocumentSegment.keywords, String).ilike(f"%{keyword}%")
+                keywords_condition = cast(DocumentSegment.keywords, String).ilike(f"%{escaped_keyword}%", escape="\\")
 
             query = query.where(
                 or_(
-                    DocumentSegment.content.ilike(f"%{keyword}%"),
+                    DocumentSegment.content.ilike(f"%{escaped_keyword}%", escape="\\"),
                     keywords_condition,
                 )
             )

api/controllers/console/datasets/hit_testing_base.py

@@ -1,7 +1,7 @@
 import logging
 from typing import Any
 
-from flask_restx import marshal, reqparse
+from flask_restx import marshal
 from pydantic import BaseModel, Field
 from werkzeug.exceptions import Forbidden, InternalServerError, NotFound
 
@@ -56,15 +56,10 @@ class DatasetsHitTestingBase:
         HitTestingService.hit_testing_args_check(args)
 
     @staticmethod
-    def parse_args():
-        parser = (
-            reqparse.RequestParser()
-            .add_argument("query", type=str, required=False, location="json")
-            .add_argument("attachment_ids", type=list, required=False, location="json")
-            .add_argument("retrieval_model", type=dict, required=False, location="json")
-            .add_argument("external_retrieval_model", type=dict, required=False, location="json")
-        )
-        return parser.parse_args()
+    def parse_args(payload: dict[str, Any]) -> dict[str, Any]:
+        """Validate and return hit-testing arguments from an incoming payload."""
+        hit_testing_payload = HitTestingPayload.model_validate(payload or {})
+        return hit_testing_payload.model_dump(exclude_none=True)
 
     @staticmethod
     def perform_hit_testing(dataset, args):

api/controllers/console/datasets/rag_pipeline/rag_pipeline_workflow.py

@@ -355,7 +355,7 @@ class PublishedRagPipelineRunApi(Resource):
                 pipeline=pipeline,
                 user=current_user,
                 args=args,
-                invoke_from=InvokeFrom.DEBUGGER if payload.is_preview else InvokeFrom.PUBLISHED,
+                invoke_from=InvokeFrom.DEBUGGER if payload.is_preview else InvokeFrom.PUBLISHED_PIPELINE,
                 streaming=streaming,
             )
 

api/controllers/console/files.py

@@ -1,7 +1,7 @@
 from typing import Literal
 
 from flask import request
-from flask_restx import Resource, marshal_with
+from flask_restx import Resource
 from werkzeug.exceptions import Forbidden
 
 import services
@@ -15,18 +15,21 @@ from controllers.common.errors import (
     TooManyFilesError,
     UnsupportedFileTypeError,
 )
+from controllers.common.schema import register_schema_models
 from controllers.console.wraps import (
     account_initialization_required,
     cloud_edition_billing_resource_check,
     setup_required,
 )
 from extensions.ext_database import db
-from fields.file_fields import file_fields, upload_config_fields
+from fields.file_fields import FileResponse, UploadConfig
 from libs.login import current_account_with_tenant, login_required
 from services.file_service import FileService
 
 from . import console_ns
 
+register_schema_models(console_ns, UploadConfig, FileResponse)
+
 PREVIEW_WORDS_LIMIT = 3000
 
 
@@ -35,26 +38,27 @@ class FileApi(Resource):
     @setup_required
     @login_required
     @account_initialization_required
-    @marshal_with(upload_config_fields)
+    @console_ns.response(200, "Success", console_ns.models[UploadConfig.__name__])
     def get(self):
-        return {
-            "file_size_limit": dify_config.UPLOAD_FILE_SIZE_LIMIT,
-            "batch_count_limit": dify_config.UPLOAD_FILE_BATCH_LIMIT,
-            "file_upload_limit": dify_config.BATCH_UPLOAD_LIMIT,
-            "image_file_size_limit": dify_config.UPLOAD_IMAGE_FILE_SIZE_LIMIT,
-            "video_file_size_limit": dify_config.UPLOAD_VIDEO_FILE_SIZE_LIMIT,
-            "audio_file_size_limit": dify_config.UPLOAD_AUDIO_FILE_SIZE_LIMIT,
-            "workflow_file_upload_limit": dify_config.WORKFLOW_FILE_UPLOAD_LIMIT,
-            "image_file_batch_limit": dify_config.IMAGE_FILE_BATCH_LIMIT,
-            "single_chunk_attachment_limit": dify_config.SINGLE_CHUNK_ATTACHMENT_LIMIT,
-            "attachment_image_file_size_limit": dify_config.ATTACHMENT_IMAGE_FILE_SIZE_LIMIT,
-        }, 200
+        config = UploadConfig(
+            file_size_limit=dify_config.UPLOAD_FILE_SIZE_LIMIT,
+            batch_count_limit=dify_config.UPLOAD_FILE_BATCH_LIMIT,
+            file_upload_limit=dify_config.BATCH_UPLOAD_LIMIT,
+            image_file_size_limit=dify_config.UPLOAD_IMAGE_FILE_SIZE_LIMIT,
+            video_file_size_limit=dify_config.UPLOAD_VIDEO_FILE_SIZE_LIMIT,
+            audio_file_size_limit=dify_config.UPLOAD_AUDIO_FILE_SIZE_LIMIT,
+            workflow_file_upload_limit=dify_config.WORKFLOW_FILE_UPLOAD_LIMIT,
+            image_file_batch_limit=dify_config.IMAGE_FILE_BATCH_LIMIT,
+            single_chunk_attachment_limit=dify_config.SINGLE_CHUNK_ATTACHMENT_LIMIT,
+            attachment_image_file_size_limit=dify_config.ATTACHMENT_IMAGE_FILE_SIZE_LIMIT,
+        )
+        return config.model_dump(mode="json"), 200
 
     @setup_required
     @login_required
     @account_initialization_required
-    @marshal_with(file_fields)
     @cloud_edition_billing_resource_check("documents")
+    @console_ns.response(201, "File uploaded successfully", console_ns.models[FileResponse.__name__])
     def post(self):
         current_user, _ = current_account_with_tenant()
         source_str = request.form.get("source")
@@ -90,7 +94,8 @@ class FileApi(Resource):
         except services.errors.file.BlockedFileExtensionError as blocked_extension_error:
             raise BlockedFileExtensionError(blocked_extension_error.description)
 
-        return upload_file, 201
+        response = FileResponse.model_validate(upload_file, from_attributes=True)
+        return response.model_dump(mode="json"), 201
 
 
 @console_ns.route("/files/<uuid:file_id>/preview")

api/controllers/console/remote_files.py

@@ -1,7 +1,7 @@
 import urllib.parse
 
 import httpx
-from flask_restx import Resource, marshal_with
+from flask_restx import Resource
 from pydantic import BaseModel, Field
 
 import services
@@ -11,19 +11,22 @@ from controllers.common.errors import (
     RemoteFileUploadError,
     UnsupportedFileTypeError,
 )
+from controllers.common.schema import register_schema_models
 from core.file import helpers as file_helpers
 from core.helper import ssrf_proxy
 from extensions.ext_database import db
-from fields.file_fields import file_fields_with_signed_url, remote_file_info_fields
+from fields.file_fields import FileWithSignedUrl, RemoteFileInfo
 from libs.login import current_account_with_tenant
 from services.file_service import FileService
 
 from . import console_ns
 
+register_schema_models(console_ns, RemoteFileInfo, FileWithSignedUrl)
+
 
 @console_ns.route("/remote-files/<path:url>")
 class RemoteFileInfoApi(Resource):
-    @marshal_with(remote_file_info_fields)
+    @console_ns.response(200, "Remote file info", console_ns.models[RemoteFileInfo.__name__])
     def get(self, url):
         decoded_url = urllib.parse.unquote(url)
         resp = ssrf_proxy.head(decoded_url)
@@ -31,10 +34,11 @@ class RemoteFileInfoApi(Resource):
             # failed back to get method
             resp = ssrf_proxy.get(decoded_url, timeout=3)
         resp.raise_for_status()
-        return {
-            "file_type": resp.headers.get("Content-Type", "application/octet-stream"),
-            "file_length": int(resp.headers.get("Content-Length", 0)),
-        }
+        info = RemoteFileInfo(
+            file_type=resp.headers.get("Content-Type", "application/octet-stream"),
+            file_length=int(resp.headers.get("Content-Length", 0)),
+        )
+        return info.model_dump(mode="json")
 
 
 class RemoteFileUploadPayload(BaseModel):
@@ -50,7 +54,7 @@ console_ns.schema_model(
 @console_ns.route("/remote-files/upload")
 class RemoteFileUploadApi(Resource):
     @console_ns.expect(console_ns.models[RemoteFileUploadPayload.__name__])
-    @marshal_with(file_fields_with_signed_url)
+    @console_ns.response(201, "Remote file uploaded", console_ns.models[FileWithSignedUrl.__name__])
     def post(self):
         args = RemoteFileUploadPayload.model_validate(console_ns.payload)
         url = args.url
@@ -85,13 +89,14 @@ class RemoteFileUploadApi(Resource):
         except services.errors.file.UnsupportedFileTypeError:
             raise UnsupportedFileTypeError()
 
-        return {
-            "id": upload_file.id,
-            "name": upload_file.name,
-            "size": upload_file.size,
-            "extension": upload_file.extension,
-            "url": file_helpers.get_signed_file_url(upload_file_id=upload_file.id),
-            "mime_type": upload_file.mime_type,
-            "created_by": upload_file.created_by,
-            "created_at": upload_file.created_at,
-        }, 201
+        payload = FileWithSignedUrl(
+            id=upload_file.id,
+            name=upload_file.name,
+            size=upload_file.size,
+            extension=upload_file.extension,
+            url=file_helpers.get_signed_file_url(upload_file_id=upload_file.id),
+            mime_type=upload_file.mime_type,
+            created_by=upload_file.created_by,
+            created_at=int(upload_file.created_at.timestamp()),
+        )
+        return payload.model_dump(mode="json"), 201

api/controllers/console/workspace/account.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 from datetime import datetime
 from typing import Literal
 
@@ -99,7 +101,7 @@ class AccountPasswordPayload(BaseModel):
     repeat_new_password: str
 
     @model_validator(mode="after")
-    def check_passwords_match(self) -> "AccountPasswordPayload":
+    def check_passwords_match(self) -> AccountPasswordPayload:
         if self.new_password != self.repeat_new_password:
             raise RepeatPasswordNotMatchError()
         return self

api/controllers/files/upload.py

@@ -4,18 +4,18 @@ from flask import request
 from flask_restx import Resource
 from flask_restx.api import HTTPStatus
 from pydantic import BaseModel, Field
-from werkzeug.datastructures import FileStorage
 from werkzeug.exceptions import Forbidden
 
 import services
 from core.file.helpers import verify_plugin_file_signature
 from core.tools.tool_file_manager import ToolFileManager
-from fields.file_fields import build_file_model
+from fields.file_fields import FileResponse
 
 from ..common.errors import (
     FileTooLargeError,
     UnsupportedFileTypeError,
 )
+from ..common.schema import register_schema_models
 from ..console.wraps import setup_required
 from ..files import files_ns
 from ..inner_api.plugin.wraps import get_user
@@ -35,6 +35,8 @@ files_ns.schema_model(
     PluginUploadQuery.__name__, PluginUploadQuery.model_json_schema(ref_template=DEFAULT_REF_TEMPLATE_SWAGGER_2_0)
 )
 
+register_schema_models(files_ns, FileResponse)
+
 
 @files_ns.route("/upload/for-plugin")
 class PluginUploadFileApi(Resource):
@@ -51,7 +53,7 @@ class PluginUploadFileApi(Resource):
             415: "Unsupported file type",
         }
     )
-    @files_ns.marshal_with(build_file_model(files_ns), code=HTTPStatus.CREATED)
+    @files_ns.response(HTTPStatus.CREATED, "File uploaded", files_ns.models[FileResponse.__name__])
     def post(self):
         """Upload a file for plugin usage.
 
@@ -69,7 +71,7 @@ class PluginUploadFileApi(Resource):
         """
         args = PluginUploadQuery.model_validate(request.args.to_dict(flat=True))  # type: ignore
 
-        file: FileStorage | None = request.files.get("file")
+        file = request.files.get("file")
         if file is None:
             raise Forbidden("File is required.")
 
@@ -80,8 +82,8 @@ class PluginUploadFileApi(Resource):
         user_id = args.user_id
         user = get_user(tenant_id, user_id)
 
-        filename: str | None = file.filename
-        mimetype: str | None = file.mimetype
+        filename = file.filename
+        mimetype = file.mimetype
 
         if not filename or not mimetype:
             raise Forbidden("Invalid request.")
@@ -111,22 +113,22 @@ class PluginUploadFileApi(Resource):
             preview_url = ToolFileManager.sign_file(tool_file_id=tool_file.id, extension=extension)
 
             # Create a dictionary with all the necessary attributes
-            result = {
-                "id": tool_file.id,
-                "user_id": tool_file.user_id,
-                "tenant_id": tool_file.tenant_id,
-                "conversation_id": tool_file.conversation_id,
-                "file_key": tool_file.file_key,
-                "mimetype": tool_file.mimetype,
-                "original_url": tool_file.original_url,
-                "name": tool_file.name,
-                "size": tool_file.size,
-                "mime_type": mimetype,
-                "extension": extension,
-                "preview_url": preview_url,
-            }
+            result = FileResponse(
+                id=tool_file.id,
+                name=tool_file.name,
+                size=tool_file.size,
+                extension=extension,
+                mime_type=mimetype,
+                preview_url=preview_url,
+                source_url=tool_file.original_url,
+                original_url=tool_file.original_url,
+                user_id=tool_file.user_id,
+                tenant_id=tool_file.tenant_id,
+                conversation_id=tool_file.conversation_id,
+                file_key=tool_file.file_key,
+            )
 
-            return result, 201
+            return result.model_dump(mode="json"), 201
         except services.errors.file.FileTooLargeError as file_too_large_error:
             raise FileTooLargeError(file_too_large_error.description)
         except services.errors.file.UnsupportedFileTypeError:

api/controllers/service_api/app/file.py

@@ -10,13 +10,16 @@ from controllers.common.errors import (
     TooManyFilesError,
     UnsupportedFileTypeError,
 )
+from controllers.common.schema import register_schema_models
 from controllers.service_api import service_api_ns
 from controllers.service_api.wraps import FetchUserArg, WhereisUserArg, validate_app_token
 from extensions.ext_database import db
-from fields.file_fields import build_file_model
+from fields.file_fields import FileResponse
 from models import App, EndUser
 from services.file_service import FileService
 
+register_schema_models(service_api_ns, FileResponse)
+
 
 @service_api_ns.route("/files/upload")
 class FileApi(Resource):
@@ -31,8 +34,8 @@ class FileApi(Resource):
             415: "Unsupported file type",
         }
     )
-    @validate_app_token(fetch_user_arg=FetchUserArg(fetch_from=WhereisUserArg.FORM))
-    @service_api_ns.marshal_with(build_file_model(service_api_ns), code=HTTPStatus.CREATED)
+    @validate_app_token(fetch_user_arg=FetchUserArg(fetch_from=WhereisUserArg.FORM))  # type: ignore
+    @service_api_ns.response(HTTPStatus.CREATED, "File uploaded", service_api_ns.models[FileResponse.__name__])
     def post(self, app_model: App, end_user: EndUser):
         """Upload a file for use in conversations.
 
@@ -64,4 +67,5 @@ class FileApi(Resource):
         except services.errors.file.UnsupportedFileTypeError:
             raise UnsupportedFileTypeError()
 
-        return upload_file, 201
+        response = FileResponse.model_validate(upload_file, from_attributes=True)
+        return response.model_dump(mode="json"), 201

api/controllers/service_api/dataset/hit_testing.py

@@ -24,7 +24,7 @@ class HitTestingApi(DatasetApiResource, DatasetsHitTestingBase):
         dataset_id_str = str(dataset_id)
 
         dataset = self.get_and_validate_dataset(dataset_id_str)
-        args = self.parse_args()
+        args = self.parse_args(service_api_ns.payload)
         self.hit_testing_args_check(args)
 
         return self.perform_hit_testing(dataset, args)

api/controllers/service_api/dataset/rag_pipeline/rag_pipeline_workflow.py

@@ -174,7 +174,7 @@ class PipelineRunApi(DatasetApiResource):
                 pipeline=pipeline,
                 user=current_user,
                 args=payload.model_dump(),
-                invoke_from=InvokeFrom.PUBLISHED if payload.is_published else InvokeFrom.DEBUGGER,
+                invoke_from=InvokeFrom.PUBLISHED_PIPELINE if payload.is_published else InvokeFrom.DEBUGGER,
                 streaming=payload.response_mode == "streaming",
             )
 

api/controllers/web/conversation.py

@@ -1,9 +1,11 @@
-from flask_restx import reqparse
-from flask_restx.inputs import int_range
-from pydantic import TypeAdapter
+from typing import Literal
+
+from flask import request
+from pydantic import BaseModel, Field, TypeAdapter, field_validator, model_validator
 from sqlalchemy.orm import Session
 from werkzeug.exceptions import NotFound
 
+from controllers.common.schema import register_schema_models
 from controllers.web import web_ns
 from controllers.web.error import NotChatAppError
 from controllers.web.wraps import WebApiResource
@@ -21,6 +23,35 @@ from services.errors.conversation import ConversationNotExistsError, LastConvers
 from services.web_conversation_service import WebConversationService
 
 
+class ConversationListQuery(BaseModel):
+    last_id: str | None = None
+    limit: int = Field(default=20, ge=1, le=100)
+    pinned: bool | None = None
+    sort_by: Literal["created_at", "-created_at", "updated_at", "-updated_at"] = "-updated_at"
+
+    @field_validator("last_id")
+    @classmethod
+    def validate_last_id(cls, value: str | None) -> str | None:
+        if value is None:
+            return value
+        return uuid_value(value)
+
+
+class ConversationRenamePayload(BaseModel):
+    name: str | None = None
+    auto_generate: bool = False
+
+    @model_validator(mode="after")
+    def validate_name_requirement(self):
+        if not self.auto_generate:
+            if self.name is None or not self.name.strip():
+                raise ValueError("name is required when auto_generate is false")
+        return self
+
+
+register_schema_models(web_ns, ConversationListQuery, ConversationRenamePayload)
+
+
 @web_ns.route("/conversations")
 class ConversationListApi(WebApiResource):
     @web_ns.doc("Get Conversation List")
@@ -64,25 +95,8 @@ class ConversationListApi(WebApiResource):
         if app_mode not in {AppMode.CHAT, AppMode.AGENT_CHAT, AppMode.ADVANCED_CHAT}:
             raise NotChatAppError()
 
-        parser = (
-            reqparse.RequestParser()
-            .add_argument("last_id", type=uuid_value, location="args")
-            .add_argument("limit", type=int_range(1, 100), required=False, default=20, location="args")
-            .add_argument("pinned", type=str, choices=["true", "false", None], location="args")
-            .add_argument(
-                "sort_by",
-                type=str,
-                choices=["created_at", "-created_at", "updated_at", "-updated_at"],
-                required=False,
-                default="-updated_at",
-                location="args",
-            )
-        )
-        args = parser.parse_args()
-
-        pinned = None
-        if "pinned" in args and args["pinned"] is not None:
-            pinned = args["pinned"] == "true"
+        raw_args = request.args.to_dict()
+        query = ConversationListQuery.model_validate(raw_args)
 
         try:
             with Session(db.engine) as session:
@@ -90,11 +104,11 @@ class ConversationListApi(WebApiResource):
                     session=session,
                     app_model=app_model,
                     user=end_user,
-                    last_id=args["last_id"],
-                    limit=args["limit"],
+                    last_id=query.last_id,
+                    limit=query.limit,
                     invoke_from=InvokeFrom.WEB_APP,
-                    pinned=pinned,
-                    sort_by=args["sort_by"],
+                    pinned=query.pinned,
+                    sort_by=query.sort_by,
                 )
                 adapter = TypeAdapter(SimpleConversation)
                 conversations = [adapter.validate_python(item, from_attributes=True) for item in pagination.data]
@@ -168,16 +182,11 @@ class ConversationRenameApi(WebApiResource):
 
         conversation_id = str(c_id)
 
-        parser = (
-            reqparse.RequestParser()
-            .add_argument("name", type=str, required=False, location="json")
-            .add_argument("auto_generate", type=bool, required=False, default=False, location="json")
-        )
-        args = parser.parse_args()
+        payload = ConversationRenamePayload.model_validate(web_ns.payload or {})
 
         try:
             conversation = ConversationService.rename(
-                app_model, conversation_id, end_user, args["name"], args["auto_generate"]
+                app_model, conversation_id, end_user, payload.name, payload.auto_generate
             )
             return (
                 TypeAdapter(SimpleConversation)

api/controllers/web/files.py

@@ -1,5 +1,4 @@
 from flask import request
-from flask_restx import marshal_with
 
 import services
 from controllers.common.errors import (
@@ -9,12 +8,15 @@ from controllers.common.errors import (
     TooManyFilesError,
     UnsupportedFileTypeError,
 )
+from controllers.common.schema import register_schema_models
 from controllers.web import web_ns
 from controllers.web.wraps import WebApiResource
 from extensions.ext_database import db
-from fields.file_fields import build_file_model
+from fields.file_fields import FileResponse
 from services.file_service import FileService
 
+register_schema_models(web_ns, FileResponse)
+
 
 @web_ns.route("/files/upload")
 class FileApi(WebApiResource):
@@ -28,7 +30,7 @@ class FileApi(WebApiResource):
             415: "Unsupported file type",
         }
     )
-    @marshal_with(build_file_model(web_ns))
+    @web_ns.response(201, "File uploaded successfully", web_ns.models[FileResponse.__name__])
     def post(self, app_model, end_user):
         """Upload a file for use in web applications.
 
@@ -81,4 +83,5 @@ class FileApi(WebApiResource):
         except services.errors.file.UnsupportedFileTypeError:
             raise UnsupportedFileTypeError()
 
-        return upload_file, 201
+        response = FileResponse.model_validate(upload_file, from_attributes=True)
+        return response.model_dump(mode="json"), 201

api/controllers/web/remote_files.py

@@ -1,7 +1,6 @@
 import urllib.parse
 
 import httpx
-from flask_restx import marshal_with
 from pydantic import BaseModel, Field, HttpUrl
 
 import services
@@ -14,7 +13,7 @@ from controllers.common.errors import (
 from core.file import helpers as file_helpers
 from core.helper import ssrf_proxy
 from extensions.ext_database import db
-from fields.file_fields import build_file_with_signed_url_model, build_remote_file_info_model
+from fields.file_fields import FileWithSignedUrl, RemoteFileInfo
 from services.file_service import FileService
 
 from ..common.schema import register_schema_models
@@ -26,7 +25,7 @@ class RemoteFileUploadPayload(BaseModel):
     url: HttpUrl = Field(description="Remote file URL")
 
 
-register_schema_models(web_ns, RemoteFileUploadPayload)
+register_schema_models(web_ns, RemoteFileUploadPayload, RemoteFileInfo, FileWithSignedUrl)
 
 
 @web_ns.route("/remote-files/<path:url>")
@@ -41,7 +40,7 @@ class RemoteFileInfoApi(WebApiResource):
             500: "Failed to fetch remote file",
         }
     )
-    @marshal_with(build_remote_file_info_model(web_ns))
+    @web_ns.response(200, "Remote file info", web_ns.models[RemoteFileInfo.__name__])
     def get(self, app_model, end_user, url):
         """Get information about a remote file.
 
@@ -65,10 +64,11 @@ class RemoteFileInfoApi(WebApiResource):
             # failed back to get method
             resp = ssrf_proxy.get(decoded_url, timeout=3)
         resp.raise_for_status()
-        return {
-            "file_type": resp.headers.get("Content-Type", "application/octet-stream"),
-            "file_length": int(resp.headers.get("Content-Length", -1)),
-        }
+        info = RemoteFileInfo(
+            file_type=resp.headers.get("Content-Type", "application/octet-stream"),
+            file_length=int(resp.headers.get("Content-Length", -1)),
+        )
+        return info.model_dump(mode="json")
 
 
 @web_ns.route("/remote-files/upload")
@@ -84,7 +84,7 @@ class RemoteFileUploadApi(WebApiResource):
             500: "Failed to fetch remote file",
         }
     )
-    @marshal_with(build_file_with_signed_url_model(web_ns))
+    @web_ns.response(201, "Remote file uploaded", web_ns.models[FileWithSignedUrl.__name__])
     def post(self, app_model, end_user):
         """Upload a file from a remote URL.
 
@@ -139,13 +139,14 @@ class RemoteFileUploadApi(WebApiResource):
         except services.errors.file.UnsupportedFileTypeError:
             raise UnsupportedFileTypeError
 
-        return {
-            "id": upload_file.id,
-            "name": upload_file.name,
-            "size": upload_file.size,
-            "extension": upload_file.extension,
-            "url": file_helpers.get_signed_file_url(upload_file_id=upload_file.id),
-            "mime_type": upload_file.mime_type,
-            "created_by": upload_file.created_by,
-            "created_at": upload_file.created_at,
-        }, 201
+        payload1 = FileWithSignedUrl(
+            id=upload_file.id,
+            name=upload_file.name,
+            size=upload_file.size,
+            extension=upload_file.extension,
+            url=file_helpers.get_signed_file_url(upload_file_id=upload_file.id),
+            mime_type=upload_file.mime_type,
+            created_by=upload_file.created_by,
+            created_at=int(upload_file.created_at.timestamp()),
+        )
+        return payload1.model_dump(mode="json"), 201

api/controllers/web/saved_message.py

@@ -1,18 +1,30 @@
-from flask_restx import reqparse
-from flask_restx.inputs import int_range
-from pydantic import TypeAdapter
+from flask import request
+from pydantic import BaseModel, Field, TypeAdapter
 from werkzeug.exceptions import NotFound
 
+from controllers.common.schema import register_schema_models
 from controllers.web import web_ns
 from controllers.web.error import NotCompletionAppError
 from controllers.web.wraps import WebApiResource
 from fields.conversation_fields import ResultResponse
 from fields.message_fields import SavedMessageInfiniteScrollPagination, SavedMessageItem
-from libs.helper import uuid_value
+from libs.helper import UUIDStrOrEmpty
 from services.errors.message import MessageNotExistsError
 from services.saved_message_service import SavedMessageService
 
 
+class SavedMessageListQuery(BaseModel):
+    last_id: UUIDStrOrEmpty | None = None
+    limit: int = Field(default=20, ge=1, le=100)
+
+
+class SavedMessageCreatePayload(BaseModel):
+    message_id: UUIDStrOrEmpty
+
+
+register_schema_models(web_ns, SavedMessageListQuery, SavedMessageCreatePayload)
+
+
 @web_ns.route("/saved-messages")
 class SavedMessageListApi(WebApiResource):
     @web_ns.doc("Get Saved Messages")
@@ -42,14 +54,10 @@ class SavedMessageListApi(WebApiResource):
         if app_model.mode != "completion":
             raise NotCompletionAppError()
 
-        parser = (
-            reqparse.RequestParser()
-            .add_argument("last_id", type=uuid_value, location="args")
-            .add_argument("limit", type=int_range(1, 100), required=False, default=20, location="args")
-        )
-        args = parser.parse_args()
+        raw_args = request.args.to_dict()
+        query = SavedMessageListQuery.model_validate(raw_args)
 
-        pagination = SavedMessageService.pagination_by_last_id(app_model, end_user, args["last_id"], args["limit"])
+        pagination = SavedMessageService.pagination_by_last_id(app_model, end_user, query.last_id, query.limit)
         adapter = TypeAdapter(SavedMessageItem)
         items = [adapter.validate_python(message, from_attributes=True) for message in pagination.data]
         return SavedMessageInfiniteScrollPagination(
@@ -79,11 +87,10 @@ class SavedMessageListApi(WebApiResource):
         if app_model.mode != "completion":
             raise NotCompletionAppError()
 
-        parser = reqparse.RequestParser().add_argument("message_id", type=uuid_value, required=True, location="json")
-        args = parser.parse_args()
+        payload = SavedMessageCreatePayload.model_validate(web_ns.payload or {})
 
         try:
-            SavedMessageService.save(app_model, end_user, args["message_id"])
+            SavedMessageService.save(app_model, end_user, payload.message_id)
         except MessageNotExistsError:
             raise NotFound("Message Not Exists.")
 

api/core/app/apps/advanced_chat/app_runner.py

@@ -20,6 +20,8 @@ from core.app.entities.queue_entities import (
     QueueTextChunkEvent,
 )
 from core.app.features.annotation_reply.annotation_reply import AnnotationReplyFeature
+from core.app.layers.conversation_variable_persist_layer import ConversationVariablePersistenceLayer
+from core.db.session_factory import session_factory
 from core.moderation.base import ModerationError
 from core.moderation.input_moderation import InputModeration
 from core.variables.variables import VariableUnion
@@ -40,6 +42,7 @@ from models import Workflow
 from models.enums import UserFrom
 from models.model import App, Conversation, Message, MessageAnnotation
 from models.workflow import ConversationVariable
+from services.conversation_variable_updater import ConversationVariableUpdater
 
 logger = logging.getLogger(__name__)
 
@@ -200,6 +203,10 @@ class AdvancedChatAppRunner(WorkflowBasedAppRunner):
         )
 
         workflow_entry.graph_engine.layer(persistence_layer)
+        conversation_variable_layer = ConversationVariablePersistenceLayer(
+            ConversationVariableUpdater(session_factory.get_session_maker())
+        )
+        workflow_entry.graph_engine.layer(conversation_variable_layer)
         for layer in self._graph_engine_layers:
             workflow_entry.graph_engine.layer(layer)
 

api/core/app/apps/pipeline/pipeline_generator.py

@@ -130,7 +130,7 @@ class PipelineGenerator(BaseAppGenerator):
             pipeline=pipeline, workflow=workflow, start_node_id=start_node_id
         )
         documents: list[Document] = []
-        if invoke_from == InvokeFrom.PUBLISHED and not is_retry and not args.get("original_document_id"):
+        if invoke_from == InvokeFrom.PUBLISHED_PIPELINE and not is_retry and not args.get("original_document_id"):
             from services.dataset_service import DocumentService
 
             for datasource_info in datasource_info_list:
@@ -156,7 +156,7 @@ class PipelineGenerator(BaseAppGenerator):
         for i, datasource_info in enumerate(datasource_info_list):
             workflow_run_id = str(uuid.uuid4())
             document_id = args.get("original_document_id") or None
-            if invoke_from == InvokeFrom.PUBLISHED and not is_retry:
+            if invoke_from == InvokeFrom.PUBLISHED_PIPELINE and not is_retry:
                 document_id = document_id or documents[i].id
                 document_pipeline_execution_log = DocumentPipelineExecutionLog(
                     document_id=document_id,

api/core/app/entities/app_invoke_entities.py

@@ -42,7 +42,8 @@ class InvokeFrom(StrEnum):
     # DEBUGGER indicates that this invocation is from
     # the workflow (or chatflow) edit page.
     DEBUGGER = "debugger"
-    PUBLISHED = "published"
+    # PUBLISHED_PIPELINE indicates that this invocation runs a published RAG pipeline workflow.
+    PUBLISHED_PIPELINE = "published"
 
     # VALIDATION indicates that this invocation is from validation.
     VALIDATION = "validation"

api/core/app/features/annotation_reply/annotation_reply.py

@@ -75,7 +75,7 @@ class AnnotationReplyFeature:
                     AppAnnotationService.add_annotation_history(
                         annotation.id,
                         app_record.id,
-                        annotation.question,
+                        annotation.question_text,
                         annotation.content,
                         query,
                         user_id,

api/core/app/layers/conversation_variable_persist_layer.py

@@ -0,0 +1,60 @@
+import logging
+
+from core.variables import Variable
+from core.workflow.constants import CONVERSATION_VARIABLE_NODE_ID
+from core.workflow.conversation_variable_updater import ConversationVariableUpdater
+from core.workflow.enums import NodeType
+from core.workflow.graph_engine.layers.base import GraphEngineLayer
+from core.workflow.graph_events import GraphEngineEvent, NodeRunSucceededEvent
+from core.workflow.nodes.variable_assigner.common import helpers as common_helpers
+
+logger = logging.getLogger(__name__)
+
+
+class ConversationVariablePersistenceLayer(GraphEngineLayer):
+    def __init__(self, conversation_variable_updater: ConversationVariableUpdater) -> None:
+        super().__init__()
+        self._conversation_variable_updater = conversation_variable_updater
+
+    def on_graph_start(self) -> None:
+        pass
+
+    def on_event(self, event: GraphEngineEvent) -> None:
+        if not isinstance(event, NodeRunSucceededEvent):
+            return
+        if event.node_type != NodeType.VARIABLE_ASSIGNER:
+            return
+        if self.graph_runtime_state is None:
+            return
+
+        updated_variables = common_helpers.get_updated_variables(event.node_run_result.process_data) or []
+        if not updated_variables:
+            return
+
+        conversation_id = self.graph_runtime_state.system_variable.conversation_id
+        if conversation_id is None:
+            return
+
+        updated_any = False
+        for item in updated_variables:
+            selector = item.selector
+            if len(selector) < 2:
+                logger.warning("Conversation variable selector invalid. selector=%s", selector)
+                continue
+            if selector[0] != CONVERSATION_VARIABLE_NODE_ID:
+                continue
+            variable = self.graph_runtime_state.variable_pool.get(selector)
+            if not isinstance(variable, Variable):
+                logger.warning(
+                    "Conversation variable not found in variable pool. selector=%s",
+                    selector,
+                )
+                continue
+            self._conversation_variable_updater.update(conversation_id=conversation_id, variable=variable)
+            updated_any = True
+
+        if updated_any:
+            self._conversation_variable_updater.flush()
+
+    def on_graph_end(self, error: Exception | None) -> None:
+        pass

api/core/app/layers/pause_state_persist_layer.py

@@ -66,6 +66,7 @@ class PauseStatePersistenceLayer(GraphEngineLayer):
         """
         if isinstance(session_factory, Engine):
             session_factory = sessionmaker(session_factory)
+        super().__init__()
         self._session_maker = session_factory
         self._state_owner_user_id = state_owner_user_id
         self._generate_entity = generate_entity
@@ -98,8 +99,6 @@ class PauseStatePersistenceLayer(GraphEngineLayer):
         if not isinstance(event, GraphRunPausedEvent):
             return
 
-        assert self.graph_runtime_state is not None
-
         entity_wrapper: _GenerateEntityUnion
         if isinstance(self._generate_entity, WorkflowAppGenerateEntity):
             entity_wrapper = _WorkflowGenerateEntityWrapper(entity=self._generate_entity)

api/core/app/layers/trigger_post_layer.py

@@ -33,6 +33,7 @@ class TriggerPostLayer(GraphEngineLayer):
         trigger_log_id: str,
         session_maker: sessionmaker[Session],
     ):
+        super().__init__()
         self.trigger_log_id = trigger_log_id
         self.start_time = start_time
         self.cfs_plan_scheduler_entity = cfs_plan_scheduler_entity
@@ -57,10 +58,6 @@ class TriggerPostLayer(GraphEngineLayer):
                 elapsed_time = (datetime.now(UTC) - self.start_time).total_seconds()
 
                 # Extract relevant data from result
-                if not self.graph_runtime_state:
-                    logger.exception("Graph runtime state is not set")
-                    return
-
                 outputs = self.graph_runtime_state.outputs
 
                 # BASICLY, workflow_execution_id is the same as workflow_run_id

api/core/datasource/__base/datasource_plugin.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 from abc import ABC, abstractmethod
 
 from configs import dify_config
@@ -30,7 +32,7 @@ class DatasourcePlugin(ABC):
         """
         return DatasourceProviderType.LOCAL_FILE
 
-    def fork_datasource_runtime(self, runtime: DatasourceRuntime) -> "DatasourcePlugin":
+    def fork_datasource_runtime(self, runtime: DatasourceRuntime) -> DatasourcePlugin:
         return self.__class__(
             entity=self.entity.model_copy(),
             runtime=runtime,

api/core/datasource/entities/datasource_entities.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 import enum
 from enum import StrEnum
 from typing import Any
@@ -31,7 +33,7 @@ class DatasourceProviderType(enum.StrEnum):
     ONLINE_DRIVE = "online_drive"
 
     @classmethod
-    def value_of(cls, value: str) -> "DatasourceProviderType":
+    def value_of(cls, value: str) -> DatasourceProviderType:
         """
         Get value of given mode.
 
@@ -81,7 +83,7 @@ class DatasourceParameter(PluginParameter):
         typ: DatasourceParameterType,
         required: bool,
         options: list[str] | None = None,
-    ) -> "DatasourceParameter":
+    ) -> DatasourceParameter:
         """
         get a simple datasource parameter
 
@@ -187,14 +189,14 @@ class DatasourceInvokeMeta(BaseModel):
     tool_config: dict | None = None
 
     @classmethod
-    def empty(cls) -> "DatasourceInvokeMeta":
+    def empty(cls) -> DatasourceInvokeMeta:
         """
         Get an empty instance of DatasourceInvokeMeta
         """
         return cls(time_cost=0.0, error=None, tool_config={})
 
     @classmethod
-    def error_instance(cls, error: str) -> "DatasourceInvokeMeta":
+    def error_instance(cls, error: str) -> DatasourceInvokeMeta:
         """
         Get an instance of DatasourceInvokeMeta with error
         """

api/core/db/session_factory.py

@@ -1,7 +1,7 @@
 from sqlalchemy import Engine
 from sqlalchemy.orm import Session, sessionmaker
 
-_session_maker: sessionmaker | None = None
+_session_maker: sessionmaker[Session] | None = None
 
 
 def configure_session_factory(engine: Engine, expire_on_commit: bool = False):
@@ -10,7 +10,7 @@ def configure_session_factory(engine: Engine, expire_on_commit: bool = False):
     _session_maker = sessionmaker(bind=engine, expire_on_commit=expire_on_commit)
 
 
-def get_session_maker() -> sessionmaker:
+def get_session_maker() -> sessionmaker[Session]:
     if _session_maker is None:
         raise RuntimeError("Session factory not configured. Call configure_session_factory() first.")
     return _session_maker
@@ -27,7 +27,7 @@ class SessionFactory:
         configure_session_factory(engine, expire_on_commit)
 
     @staticmethod
-    def get_session_maker() -> sessionmaker:
+    def get_session_maker() -> sessionmaker[Session]:
         return get_session_maker()
 
     @staticmethod

api/core/entities/mcp_provider.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 import json
 from datetime import datetime
 from enum import StrEnum
@@ -75,7 +77,7 @@ class MCPProviderEntity(BaseModel):
     updated_at: datetime
 
     @classmethod
-    def from_db_model(cls, db_provider: "MCPToolProvider") -> "MCPProviderEntity":
+    def from_db_model(cls, db_provider: MCPToolProvider) -> MCPProviderEntity:
         """Create entity from database model with decryption"""
 
         return cls(

api/core/entities/provider_entities.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 from enum import StrEnum, auto
 from typing import Union
 
@@ -178,7 +180,7 @@ class BasicProviderConfig(BaseModel):
         TOOLS_SELECTOR = CommonParameterType.TOOLS_SELECTOR
 
         @classmethod
-        def value_of(cls, value: str) -> "ProviderConfig.Type":
+        def value_of(cls, value: str) -> ProviderConfig.Type:
             """
             Get value of given mode.
 

api/core/file/helpers.py

@@ -8,8 +8,9 @@ import urllib.parse
 from configs import dify_config
 
 
-def get_signed_file_url(upload_file_id: str, as_attachment=False) -> str:
-    url = f"{dify_config.FILES_URL}/files/{upload_file_id}/file-preview"
+def get_signed_file_url(upload_file_id: str, as_attachment=False, for_external: bool = True) -> str:
+    base_url = dify_config.FILES_URL if for_external else (dify_config.INTERNAL_FILES_URL or dify_config.FILES_URL)
+    url = f"{base_url}/files/{upload_file_id}/file-preview"
 
     timestamp = str(int(time.time()))
     nonce = os.urandom(16).hex()

api/core/file/models.py

@@ -112,17 +112,17 @@ class File(BaseModel):
 
         return text
 
-    def generate_url(self) -> str | None:
+    def generate_url(self, for_external: bool = True) -> str | None:
         if self.transfer_method == FileTransferMethod.REMOTE_URL:
             return self.remote_url
         elif self.transfer_method == FileTransferMethod.LOCAL_FILE:
             if self.related_id is None:
                 raise ValueError("Missing file related_id")
-            return helpers.get_signed_file_url(upload_file_id=self.related_id)
+            return helpers.get_signed_file_url(upload_file_id=self.related_id, for_external=for_external)
         elif self.transfer_method in [FileTransferMethod.TOOL_FILE, FileTransferMethod.DATASOURCE_FILE]:
             assert self.related_id is not None
             assert self.extension is not None
-            return sign_tool_file(tool_file_id=self.related_id, extension=self.extension)
+            return sign_tool_file(tool_file_id=self.related_id, extension=self.extension, for_external=for_external)
         return None
 
     def to_plugin_parameter(self) -> dict[str, Any]:
@@ -133,7 +133,7 @@ class File(BaseModel):
             "extension": self.extension,
             "size": self.size,
             "type": self.type,
-            "url": self.generate_url(),
+            "url": self.generate_url(for_external=False),
         }
 
     @model_validator(mode="after")

api/core/helper/code_executor/template_transformer.py

@@ -76,7 +76,7 @@ class TemplateTransformer(ABC):
         Post-process the result to convert scientific notation strings back to numbers
         """
 
-        def convert_scientific_notation(value):
+        def convert_scientific_notation(value: Any) -> Any:
             if isinstance(value, str):
                 # Check if the string looks like scientific notation
                 if re.match(r"^-?\d+\.?\d*e[+-]\d+$", value, re.IGNORECASE):
@@ -90,7 +90,7 @@ class TemplateTransformer(ABC):
                 return [convert_scientific_notation(v) for v in value]
             return value
 
-        return convert_scientific_notation(result)  # type: ignore[no-any-return]
+        return convert_scientific_notation(result)
 
     @classmethod
     @abstractmethod

api/core/mcp/session/base_session.py

@@ -68,13 +68,7 @@ class RequestResponder(Generic[ReceiveRequestT, SendResultT]):
         request_id: RequestId,
         request_meta: RequestParams.Meta | None,
         request: ReceiveRequestT,
-        session: """BaseSession[
-            SendRequestT,
-            SendNotificationT,
-            SendResultT,
-            ReceiveRequestT,
-            ReceiveNotificationT
-        ]""",
+        session: """BaseSession[SendRequestT, SendNotificationT, SendResultT, ReceiveRequestT, ReceiveNotificationT]""",
         on_complete: Callable[["RequestResponder[ReceiveRequestT, SendResultT]"], Any],
     ):
         self.request_id = request_id

api/core/model_runtime/entities/message_entities.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 from abc import ABC
 from collections.abc import Mapping, Sequence
 from enum import StrEnum, auto
@@ -17,7 +19,7 @@ class PromptMessageRole(StrEnum):
     TOOL = auto()
 
     @classmethod
-    def value_of(cls, value: str) -> "PromptMessageRole":
+    def value_of(cls, value: str) -> PromptMessageRole:
         """
         Get value of given mode.
 

api/core/model_runtime/entities/model_entities.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 from decimal import Decimal
 from enum import StrEnum, auto
 from typing import Any
@@ -20,7 +22,7 @@ class ModelType(StrEnum):
     TTS = auto()
 
     @classmethod
-    def value_of(cls, origin_model_type: str) -> "ModelType":
+    def value_of(cls, origin_model_type: str) -> ModelType:
         """
         Get model type from origin model type.
 
@@ -103,7 +105,7 @@ class DefaultParameterName(StrEnum):
     JSON_SCHEMA = auto()
 
     @classmethod
-    def value_of(cls, value: Any) -> "DefaultParameterName":
+    def value_of(cls, value: Any) -> DefaultParameterName:
         """
         Get parameter name from value.
 

api/core/model_runtime/model_providers/model_provider_factory.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 import hashlib
 import logging
 from collections.abc import Sequence
@@ -38,7 +40,7 @@ class ModelProviderFactory:
         plugin_providers = self.get_plugin_model_providers()
         return [provider.declaration for provider in plugin_providers]
 
-    def get_plugin_model_providers(self) -> Sequence["PluginModelProviderEntity"]:
+    def get_plugin_model_providers(self) -> Sequence[PluginModelProviderEntity]:
         """
         Get all plugin model providers
         :return: list of plugin model providers
@@ -76,7 +78,7 @@ class ModelProviderFactory:
         plugin_model_provider_entity = self.get_plugin_model_provider(provider=provider)
         return plugin_model_provider_entity.declaration
 
-    def get_plugin_model_provider(self, provider: str) -> "PluginModelProviderEntity":
+    def get_plugin_model_provider(self, provider: str) -> PluginModelProviderEntity:
         """
         Get plugin model provider
         :param provider: provider name

api/core/plugin/entities/plugin_daemon.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 import enum
 from collections.abc import Mapping, Sequence
 from datetime import datetime
@@ -242,7 +244,7 @@ class CredentialType(enum.StrEnum):
         return [item.value for item in cls]
 
     @classmethod
-    def of(cls, credential_type: str) -> "CredentialType":
+    def of(cls, credential_type: str) -> CredentialType:
         type_name = credential_type.lower()
         if type_name in {"api-key", "api_key"}:
             return cls.API_KEY

api/core/rag/datasource/vdb/clickzetta/clickzetta_vector.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 import contextlib
 import json
 import logging
@@ -6,7 +8,7 @@ import re
 import threading
 import time
 import uuid
-from typing import TYPE_CHECKING, Any, Optional
+from typing import TYPE_CHECKING, Any
 
 import clickzetta  # type: ignore
 from pydantic import BaseModel, model_validator
@@ -76,7 +78,7 @@ class ClickzettaConnectionPool:
     Manages connection reuse across ClickzettaVector instances.
     """
 
-    _instance: Optional["ClickzettaConnectionPool"] = None
+    _instance: ClickzettaConnectionPool | None = None
     _lock = threading.Lock()
 
     def __init__(self):
@@ -89,7 +91,7 @@ class ClickzettaConnectionPool:
         self._start_cleanup_thread()
 
     @classmethod
-    def get_instance(cls) -> "ClickzettaConnectionPool":
+    def get_instance(cls) -> ClickzettaConnectionPool:
         """Get singleton instance of connection pool."""
         if cls._instance is None:
             with cls._lock:
@@ -104,7 +106,7 @@ class ClickzettaConnectionPool:
             f"{config.workspace}:{config.vcluster}:{config.schema_name}"
         )
 
-    def _create_connection(self, config: ClickzettaConfig) -> "Connection":
+    def _create_connection(self, config: ClickzettaConfig) -> Connection:
         """Create a new ClickZetta connection."""
         max_retries = 3
         retry_delay = 1.0
@@ -134,7 +136,7 @@ class ClickzettaConnectionPool:
 
         raise RuntimeError(f"Failed to create ClickZetta connection after {max_retries} attempts")
 
-    def _configure_connection(self, connection: "Connection"):
+    def _configure_connection(self, connection: Connection):
         """Configure connection session settings."""
         try:
             with connection.cursor() as cursor:
@@ -181,7 +183,7 @@ class ClickzettaConnectionPool:
         except Exception:
             logger.exception("Failed to configure connection, continuing with defaults")
 
-    def _is_connection_valid(self, connection: "Connection") -> bool:
+    def _is_connection_valid(self, connection: Connection) -> bool:
         """Check if connection is still valid."""
         try:
             with connection.cursor() as cursor:
@@ -190,7 +192,7 @@ class ClickzettaConnectionPool:
         except Exception:
             return False
 
-    def get_connection(self, config: ClickzettaConfig) -> "Connection":
+    def get_connection(self, config: ClickzettaConfig) -> Connection:
         """Get a connection from the pool or create a new one."""
         config_key = self._get_config_key(config)
 
@@ -221,7 +223,7 @@ class ClickzettaConnectionPool:
             # No valid connection found, create new one
             return self._create_connection(config)
 
-    def return_connection(self, config: ClickzettaConfig, connection: "Connection"):
+    def return_connection(self, config: ClickzettaConfig, connection: Connection):
         """Return a connection to the pool."""
         config_key = self._get_config_key(config)
 
@@ -315,22 +317,22 @@ class ClickzettaVector(BaseVector):
         self._connection_pool = ClickzettaConnectionPool.get_instance()
         self._init_write_queue()
 
-    def _get_connection(self) -> "Connection":
+    def _get_connection(self) -> Connection:
         """Get a connection from the pool."""
         return self._connection_pool.get_connection(self._config)
 
-    def _return_connection(self, connection: "Connection"):
+    def _return_connection(self, connection: Connection):
         """Return a connection to the pool."""
         self._connection_pool.return_connection(self._config, connection)
 
     class ConnectionContext:
         """Context manager for borrowing and returning connections."""
 
-        def __init__(self, vector_instance: "ClickzettaVector"):
+        def __init__(self, vector_instance: ClickzettaVector):
             self.vector = vector_instance
             self.connection: Connection | None = None
 
-        def __enter__(self) -> "Connection":
+        def __enter__(self) -> Connection:
             self.connection = self.vector._get_connection()
             return self.connection
 
@@ -338,7 +340,7 @@ class ClickzettaVector(BaseVector):
             if self.connection:
                 self.vector._return_connection(self.connection)
 
-    def get_connection_context(self) -> "ClickzettaVector.ConnectionContext":
+    def get_connection_context(self) -> ClickzettaVector.ConnectionContext:
         """Get a connection context manager."""
         return self.ConnectionContext(self)
 
@@ -437,7 +439,7 @@ class ClickzettaVector(BaseVector):
         """Return the vector database type."""
         return "clickzetta"
 
-    def _ensure_connection(self) -> "Connection":
+    def _ensure_connection(self) -> Connection:
         """Get a connection from the pool."""
         return self._get_connection()
 
@@ -984,9 +986,11 @@ class ClickzettaVector(BaseVector):
 
         # No need for dataset_id filter since each dataset has its own table
 
-        # Use simple quote escaping for LIKE clause
-        escaped_query = query.replace("'", "''")
-        filter_clauses.append(f"{Field.CONTENT_KEY} LIKE '%{escaped_query}%'")
+        # Escape special characters for LIKE clause to prevent SQL injection
+        from libs.helper import escape_like_pattern
+
+        escaped_query = escape_like_pattern(query).replace("'", "''")
+        filter_clauses.append(f"{Field.CONTENT_KEY} LIKE '%{escaped_query}%' ESCAPE '\\\\'")
         where_clause = " AND ".join(filter_clauses)
 
         search_sql = f"""

api/core/rag/datasource/vdb/iris/iris_vector.py

@@ -287,11 +287,15 @@ class IrisVector(BaseVector):
                 cursor.execute(sql, (query,))
             else:
                 # Fallback to LIKE search (inefficient for large datasets)
-                query_pattern = f"%{query}%"
+                # Escape special characters for LIKE clause to prevent SQL injection
+                from libs.helper import escape_like_pattern
+
+                escaped_query = escape_like_pattern(query)
+                query_pattern = f"%{escaped_query}%"
                 sql = f"""
                     SELECT TOP {top_k} id, text, meta
                     FROM {self.schema}.{self.table_name}
-                    WHERE text LIKE ?
+                    WHERE text LIKE ? ESCAPE '\\'
                 """
                 cursor.execute(sql, (query_pattern,))
 

api/core/rag/datasource/vdb/weaviate/weaviate_vector.py

@@ -66,6 +66,8 @@ class WeaviateVector(BaseVector):
     in a Weaviate collection.
     """
 
+    _DOCUMENT_ID_PROPERTY = "document_id"
+
     def __init__(self, collection_name: str, config: WeaviateConfig, attributes: list):
         """
         Initializes the Weaviate vector store.
@@ -353,15 +355,12 @@ class WeaviateVector(BaseVector):
             return []
 
         col = self._client.collections.use(self._collection_name)
-        props = list({*self._attributes, "document_id", Field.TEXT_KEY.value})
+        props = list({*self._attributes, self._DOCUMENT_ID_PROPERTY, Field.TEXT_KEY.value})
 
         where = None
         doc_ids = kwargs.get("document_ids_filter") or []
         if doc_ids:
-            ors = [Filter.by_property("document_id").equal(x) for x in doc_ids]
-            where = ors[0]
-            for f in ors[1:]:
-                where = where | f
+            where = Filter.by_property(self._DOCUMENT_ID_PROPERTY).contains_any(doc_ids)
 
         top_k = int(kwargs.get("top_k", 4))
         score_threshold = float(kwargs.get("score_threshold") or 0.0)
@@ -408,10 +407,7 @@ class WeaviateVector(BaseVector):
         where = None
         doc_ids = kwargs.get("document_ids_filter") or []
         if doc_ids:
-            ors = [Filter.by_property("document_id").equal(x) for x in doc_ids]
-            where = ors[0]
-            for f in ors[1:]:
-                where = where | f
+            where = Filter.by_property(self._DOCUMENT_ID_PROPERTY).contains_any(doc_ids)
 
         top_k = int(kwargs.get("top_k", 4))
 

api/core/rag/docstore/dataset_docstore.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 from collections.abc import Sequence
 from typing import Any
 
@@ -22,7 +24,7 @@ class DatasetDocumentStore:
         self._document_id = document_id
 
     @classmethod
-    def from_dict(cls, config_dict: dict[str, Any]) -> "DatasetDocumentStore":
+    def from_dict(cls, config_dict: dict[str, Any]) -> DatasetDocumentStore:
         return cls(**config_dict)
 
     def to_dict(self) -> dict[str, Any]:

api/core/rag/extractor/word_extractor.py

@@ -7,10 +7,11 @@ import re
 import tempfile
 import uuid
 from urllib.parse import urlparse
-from xml.etree import ElementTree
 
 import httpx
 from docx import Document as DocxDocument
+from docx.oxml.ns import qn
+from docx.text.run import Run
 
 from configs import dify_config
 from core.helper import ssrf_proxy
@@ -229,44 +230,20 @@ class WordExtractor(BaseExtractor):
 
         image_map = self._extract_images_from_docx(doc)
 
-        hyperlinks_url = None
-        url_pattern = re.compile(r"http://[^\s+]+//|https://[^\s+]+")
-        for para in doc.paragraphs:
-            for run in para.runs:
-                if run.text and hyperlinks_url:
-                    result = f"  [{run.text}]({hyperlinks_url})  "
-                    run.text = result
-                    hyperlinks_url = None
-                if "HYPERLINK" in run.element.xml:
-                    try:
-                        xml = ElementTree.XML(run.element.xml)
-                        x_child = [c for c in xml.iter() if c is not None]
-                        for x in x_child:
-                            if x is None:
-                                continue
-                            if x.tag.endswith("instrText"):
-                                if x.text is None:
-                                    continue
-                                for i in url_pattern.findall(x.text):
-                                    hyperlinks_url = str(i)
-                    except Exception:
-                        logger.exception("Failed to parse HYPERLINK xml")
-
         def parse_paragraph(paragraph):
-            paragraph_content = []
-
-            def append_image_link(image_id, has_drawing):
+            def append_image_link(image_id, has_drawing, target_buffer):
                 """Helper to append image link from image_map based on relationship type."""
                 rel = doc.part.rels[image_id]
                 if rel.is_external:
                     if image_id in image_map and not has_drawing:
-                        paragraph_content.append(image_map[image_id])
+                        target_buffer.append(image_map[image_id])
                 else:
                     image_part = rel.target_part
                     if image_part in image_map and not has_drawing:
-                        paragraph_content.append(image_map[image_part])
+                        target_buffer.append(image_map[image_part])
 
-            for run in paragraph.runs:
+            def process_run(run, target_buffer):
+                # Helper to extract text and embedded images from a run element and append them to target_buffer
                 if hasattr(run.element, "tag") and isinstance(run.element.tag, str) and run.element.tag.endswith("r"):
                     # Process drawing type images
                     drawing_elements = run.element.findall(
@@ -287,13 +264,13 @@ class WordExtractor(BaseExtractor):
                                     # External image: use embed_id as key
                                     if embed_id in image_map:
                                         has_drawing = True
-                                        paragraph_content.append(image_map[embed_id])
+                                        target_buffer.append(image_map[embed_id])
                                 else:
                                     # Internal image: use target_part as key
                                     image_part = doc.part.related_parts.get(embed_id)
                                     if image_part in image_map:
                                         has_drawing = True
-                                        paragraph_content.append(image_map[image_part])
+                                        target_buffer.append(image_map[image_part])
                     # Process pict type images
                     shape_elements = run.element.findall(
                         ".//{http://schemas.openxmlformats.org/wordprocessingml/2006/main}pict"
@@ -308,7 +285,7 @@ class WordExtractor(BaseExtractor):
                                 "{http://schemas.openxmlformats.org/officeDocument/2006/relationships}id"
                             )
                             if image_id and image_id in doc.part.rels:
-                                append_image_link(image_id, has_drawing)
+                                append_image_link(image_id, has_drawing, target_buffer)
                         # Find imagedata element in VML
                         image_data = shape.find(".//{urn:schemas-microsoft-com:vml}imagedata")
                         if image_data is not None:
@@ -316,9 +293,93 @@ class WordExtractor(BaseExtractor):
                                 "{http://schemas.openxmlformats.org/officeDocument/2006/relationships}id"
                             )
                             if image_id and image_id in doc.part.rels:
-                                append_image_link(image_id, has_drawing)
+                                append_image_link(image_id, has_drawing, target_buffer)
                 if run.text.strip():
-                    paragraph_content.append(run.text.strip())
+                    target_buffer.append(run.text.strip())
+
+            def process_hyperlink(hyperlink_elem, target_buffer):
+                # Helper to extract text from a hyperlink element and append it to target_buffer
+                r_id = hyperlink_elem.get(qn("r:id"))
+
+                # Extract text from runs inside the hyperlink
+                link_text_parts = []
+                for run_elem in hyperlink_elem.findall(qn("w:r")):
+                    run = Run(run_elem, paragraph)
+                    # Hyperlink text may be split across multiple runs (e.g., with different formatting),
+                    # so collect all run texts first
+                    if run.text:
+                        link_text_parts.append(run.text)
+
+                link_text = "".join(link_text_parts).strip()
+
+                # Resolve URL
+                if r_id:
+                    try:
+                        rel = doc.part.rels.get(r_id)
+                        if rel and rel.is_external:
+                            link_text = f"[{link_text or rel.target_ref}]({rel.target_ref})"
+                    except Exception:
+                        logger.exception("Failed to resolve URL for hyperlink with r:id: %s", r_id)
+
+                if link_text:
+                    target_buffer.append(link_text)
+
+            paragraph_content = []
+            # State for legacy HYPERLINK fields
+            hyperlink_field_url = None
+            hyperlink_field_text_parts: list = []
+            is_collecting_field_text = False
+            # Iterate through paragraph elements in document order
+            for child in paragraph._element:
+                tag = child.tag
+                if tag == qn("w:r"):
+                    # Regular run
+                    run = Run(child, paragraph)
+
+                    # Check for fldChar (begin/end/separate) and instrText for legacy hyperlinks
+                    fld_chars = child.findall(qn("w:fldChar"))
+                    instr_texts = child.findall(qn("w:instrText"))
+
+                    # Handle Fields
+                    if fld_chars or instr_texts:
+                        # Process instrText to find HYPERLINK "url"
+                        for instr in instr_texts:
+                            if instr.text and "HYPERLINK" in instr.text:
+                                # Quick regex to extract URL
+                                match = re.search(r'HYPERLINK\s+"([^"]+)"', instr.text, re.IGNORECASE)
+                                if match:
+                                    hyperlink_field_url = match.group(1)
+
+                        # Process fldChar
+                        for fld_char in fld_chars:
+                            fld_char_type = fld_char.get(qn("w:fldCharType"))
+                            if fld_char_type == "begin":
+                                # Start of a field: reset legacy link state
+                                hyperlink_field_url = None
+                                hyperlink_field_text_parts = []
+                                is_collecting_field_text = False
+                            elif fld_char_type == "separate":
+                                # Separator: if we found a URL, start collecting visible text
+                                if hyperlink_field_url:
+                                    is_collecting_field_text = True
+                            elif fld_char_type == "end":
+                                # End of field
+                                if is_collecting_field_text and hyperlink_field_url:
+                                    # Create markdown link and append to main content
+                                    display_text = "".join(hyperlink_field_text_parts).strip()
+                                    if display_text:
+                                        link_md = f"[{display_text}]({hyperlink_field_url})"
+                                        paragraph_content.append(link_md)
+                                # Reset state
+                                hyperlink_field_url = None
+                                hyperlink_field_text_parts = []
+                                is_collecting_field_text = False
+
+                    # Decide where to append content
+                    target_buffer = hyperlink_field_text_parts if is_collecting_field_text else paragraph_content
+                    process_run(run, target_buffer)
+                elif tag == qn("w:hyperlink"):
+                    process_hyperlink(child, paragraph_content)
             return "".join(paragraph_content) if paragraph_content else ""
 
         paragraphs = doc.paragraphs.copy()

api/core/rag/pipeline/queue.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 import json
 from collections.abc import Sequence
 from typing import Any
@@ -16,7 +18,7 @@ class TaskWrapper(BaseModel):
         return self.model_dump_json()
 
     @classmethod
-    def deserialize(cls, serialized_data: str) -> "TaskWrapper":
+    def deserialize(cls, serialized_data: str) -> TaskWrapper:
         return cls.model_validate_json(serialized_data)
 
 

api/core/rag/retrieval/dataset_retrieval.py

@@ -1198,18 +1198,24 @@ class DatasetRetrieval:
 
         json_field = DatasetDocument.doc_metadata[metadata_name].as_string()
 
+        from libs.helper import escape_like_pattern
+
         match condition:
             case "contains":
-                filters.append(json_field.like(f"%{value}%"))
+                escaped_value = escape_like_pattern(str(value))
+                filters.append(json_field.like(f"%{escaped_value}%", escape="\\"))
 
             case "not contains":
-                filters.append(json_field.notlike(f"%{value}%"))
+                escaped_value = escape_like_pattern(str(value))
+                filters.append(json_field.notlike(f"%{escaped_value}%", escape="\\"))
 
             case "start with":
-                filters.append(json_field.like(f"{value}%"))
+                escaped_value = escape_like_pattern(str(value))
+                filters.append(json_field.like(f"{escaped_value}%", escape="\\"))
 
             case "end with":
-                filters.append(json_field.like(f"%{value}"))
+                escaped_value = escape_like_pattern(str(value))
+                filters.append(json_field.like(f"%{escaped_value}", escape="\\"))
 
             case "is" | "=":
                 if isinstance(value, str):

api/core/schemas/registry.py

@@ -1,9 +1,11 @@
+from __future__ import annotations
+
 import json
 import logging
 import threading
 from collections.abc import Mapping, MutableMapping
 from pathlib import Path
-from typing import Any, ClassVar, Optional
+from typing import Any, ClassVar
 
 
 class SchemaRegistry:
@@ -11,7 +13,7 @@ class SchemaRegistry:
 
     logger: ClassVar[logging.Logger] = logging.getLogger(__name__)
 
-    _default_instance: ClassVar[Optional["SchemaRegistry"]] = None
+    _default_instance: ClassVar[SchemaRegistry | None] = None
     _lock: ClassVar[threading.Lock] = threading.Lock()
 
     def __init__(self, base_dir: str):
@@ -20,7 +22,7 @@ class SchemaRegistry:
         self.metadata: MutableMapping[str, MutableMapping[str, Any]] = {}
 
     @classmethod
-    def default_registry(cls) -> "SchemaRegistry":
+    def default_registry(cls) -> SchemaRegistry:
         """Returns the default schema registry for builtin schemas (thread-safe singleton)"""
         if cls._default_instance is None:
             with cls._lock:

api/core/tools/__base/tool.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 from abc import ABC, abstractmethod
 from collections.abc import Generator
 from copy import deepcopy
@@ -24,7 +26,7 @@ class Tool(ABC):
         self.entity = entity
         self.runtime = runtime
 
-    def fork_tool_runtime(self, runtime: ToolRuntime) -> "Tool":
+    def fork_tool_runtime(self, runtime: ToolRuntime) -> Tool:
         """
         fork a new tool with metadata
         :return: the new tool
@@ -166,7 +168,7 @@ class Tool(ABC):
             type=ToolInvokeMessage.MessageType.IMAGE, message=ToolInvokeMessage.TextMessage(text=image)
         )
 
-    def create_file_message(self, file: "File") -> ToolInvokeMessage:
+    def create_file_message(self, file: File) -> ToolInvokeMessage:
         return ToolInvokeMessage(
             type=ToolInvokeMessage.MessageType.FILE,
             message=ToolInvokeMessage.FileMessage(),

api/core/tools/builtin_tool/tool.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 from core.model_runtime.entities.llm_entities import LLMResult
 from core.model_runtime.entities.message_entities import PromptMessage, SystemPromptMessage, UserPromptMessage
 from core.tools.__base.tool import Tool
@@ -24,7 +26,7 @@ class BuiltinTool(Tool):
         super().__init__(**kwargs)
         self.provider = provider
 
-    def fork_tool_runtime(self, runtime: ToolRuntime) -> "BuiltinTool":
+    def fork_tool_runtime(self, runtime: ToolRuntime) -> BuiltinTool:
         """
         fork a new tool with metadata
         :return: the new tool

api/core/tools/custom_tool/provider.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 from pydantic import Field
 from sqlalchemy import select
 
@@ -32,7 +34,7 @@ class ApiToolProviderController(ToolProviderController):
         self.tools = []
 
     @classmethod
-    def from_db(cls, db_provider: ApiToolProvider, auth_type: ApiProviderAuthType) -> "ApiToolProviderController":
+    def from_db(cls, db_provider: ApiToolProvider, auth_type: ApiProviderAuthType) -> ApiToolProviderController:
         credentials_schema = [
             ProviderConfig(
                 name="auth_type",

api/core/tools/entities/tool_entities.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 import base64
 import contextlib
 from collections.abc import Mapping
@@ -55,7 +57,7 @@ class ToolProviderType(StrEnum):
     MCP = auto()
 
     @classmethod
-    def value_of(cls, value: str) -> "ToolProviderType":
+    def value_of(cls, value: str) -> ToolProviderType:
         """
         Get value of given mode.
 
@@ -79,7 +81,7 @@ class ApiProviderSchemaType(StrEnum):
     OPENAI_ACTIONS = auto()
 
     @classmethod
-    def value_of(cls, value: str) -> "ApiProviderSchemaType":
+    def value_of(cls, value: str) -> ApiProviderSchemaType:
         """
         Get value of given mode.
 
@@ -102,7 +104,7 @@ class ApiProviderAuthType(StrEnum):
     API_KEY_QUERY = auto()
 
     @classmethod
-    def value_of(cls, value: str) -> "ApiProviderAuthType":
+    def value_of(cls, value: str) -> ApiProviderAuthType:
         """
         Get value of given mode.
 
@@ -307,7 +309,7 @@ class ToolParameter(PluginParameter):
         typ: ToolParameterType,
         required: bool,
         options: list[str] | None = None,
-    ) -> "ToolParameter":
+    ) -> ToolParameter:
         """
         get a simple tool parameter
 
@@ -429,14 +431,14 @@ class ToolInvokeMeta(BaseModel):
     tool_config: dict | None = None
 
     @classmethod
-    def empty(cls) -> "ToolInvokeMeta":
+    def empty(cls) -> ToolInvokeMeta:
         """
         Get an empty instance of ToolInvokeMeta
         """
         return cls(time_cost=0.0, error=None, tool_config={})
 
     @classmethod
-    def error_instance(cls, error: str) -> "ToolInvokeMeta":
+    def error_instance(cls, error: str) -> ToolInvokeMeta:
         """
         Get an instance of ToolInvokeMeta with error
         """

api/core/tools/mcp_tool/tool.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 import base64
 import json
 import logging
@@ -118,7 +120,7 @@ class MCPTool(Tool):
         for item in json_list:
             yield self.create_json_message(item)
 
-    def fork_tool_runtime(self, runtime: ToolRuntime) -> "MCPTool":
+    def fork_tool_runtime(self, runtime: ToolRuntime) -> MCPTool:
         return MCPTool(
             entity=self.entity,
             runtime=runtime,

api/core/tools/plugin_tool/tool.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 from collections.abc import Generator
 from typing import Any
 
@@ -46,7 +48,7 @@ class PluginTool(Tool):
             message_id=message_id,
         )
 
-    def fork_tool_runtime(self, runtime: ToolRuntime) -> "PluginTool":
+    def fork_tool_runtime(self, runtime: ToolRuntime) -> PluginTool:
         return PluginTool(
             entity=self.entity,
             runtime=runtime,

api/core/tools/signature.py

@@ -7,12 +7,12 @@ import time
 from configs import dify_config
 
 
-def sign_tool_file(tool_file_id: str, extension: str) -> str:
+def sign_tool_file(tool_file_id: str, extension: str, for_external: bool = True) -> str:
     """
     sign file to get a temporary url for plugin access
     """
-    # Use internal URL for plugin/tool file access in Docker environments
-    base_url = dify_config.INTERNAL_FILES_URL or dify_config.FILES_URL
+    # Use internal URL for plugin/tool file access in Docker environments, unless for_external is True
+    base_url = dify_config.FILES_URL if for_external else (dify_config.INTERNAL_FILES_URL or dify_config.FILES_URL)
     file_preview_url = f"{base_url}/files/tools/{tool_file_id}{extension}"
 
     timestamp = str(int(time.time()))

api/core/tools/workflow_as_tool/provider.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 from collections.abc import Mapping
 
 from pydantic import Field
@@ -47,7 +49,7 @@ class WorkflowToolProviderController(ToolProviderController):
         self.provider_id = provider_id
 
     @classmethod
-    def from_db(cls, db_provider: WorkflowToolProvider) -> "WorkflowToolProviderController":
+    def from_db(cls, db_provider: WorkflowToolProvider) -> WorkflowToolProviderController:
         with session_factory.create_session() as session, session.begin():
             app = session.get(App, db_provider.app_id)
             if not app:

api/core/tools/workflow_as_tool/tool.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 import json
 import logging
 from collections.abc import Generator, Mapping, Sequence
@@ -181,7 +183,7 @@ class WorkflowTool(Tool):
                             return found
         return None
 
-    def fork_tool_runtime(self, runtime: ToolRuntime) -> "WorkflowTool":
+    def fork_tool_runtime(self, runtime: ToolRuntime) -> WorkflowTool:
         """
         fork a new tool with metadata
 

api/core/variables/types.py

@@ -1,6 +1,8 @@
+from __future__ import annotations
+
 from collections.abc import Mapping
 from enum import StrEnum
-from typing import TYPE_CHECKING, Any, Optional
+from typing import TYPE_CHECKING, Any
 
 from core.file.models import File
 
@@ -52,7 +54,7 @@ class SegmentType(StrEnum):
         return self in _ARRAY_TYPES
 
     @classmethod
-    def infer_segment_type(cls, value: Any) -> Optional["SegmentType"]:
+    def infer_segment_type(cls, value: Any) -> SegmentType | None:
         """
         Attempt to infer the `SegmentType` based on the Python type of the `value` parameter.
 
@@ -173,7 +175,7 @@ class SegmentType(StrEnum):
             raise AssertionError("this statement should be unreachable.")
 
     @staticmethod
-    def cast_value(value: Any, type_: "SegmentType"):
+    def cast_value(value: Any, type_: SegmentType):
         # Cast Python's `bool` type to `int` when the runtime type requires
         # an integer or number.
         #
@@ -193,7 +195,7 @@ class SegmentType(StrEnum):
             return [int(i) for i in value]
         return value
 
-    def exposed_type(self) -> "SegmentType":
+    def exposed_type(self) -> SegmentType:
         """Returns the type exposed to the frontend.
 
         The frontend treats `INTEGER` and `FLOAT` as `NUMBER`, so these are returned as `NUMBER` here.
@@ -202,7 +204,7 @@ class SegmentType(StrEnum):
             return SegmentType.NUMBER
         return self
 
-    def element_type(self) -> "SegmentType | None":
+    def element_type(self) -> SegmentType | None:
         """Return the element type of the current segment type, or `None` if the element type is undefined.
 
         Raises:
@@ -217,7 +219,7 @@ class SegmentType(StrEnum):
         return _ARRAY_ELEMENT_TYPES_MAPPING.get(self)
 
     @staticmethod
-    def get_zero_value(t: "SegmentType"):
+    def get_zero_value(t: SegmentType):
         # Lazy import to avoid circular dependency
         from factories import variable_factory
 

api/core/workflow/README.md

@@ -64,6 +64,9 @@ engine.layer(DebugLoggingLayer(level="INFO"))
 engine.layer(ExecutionLimitsLayer(max_nodes=100))
 ```
 
+`engine.layer()` binds the read-only runtime state before execution, so layer hooks
+can assume `graph_runtime_state` is available.
+
 ### Event-Driven Architecture
 
 All node executions emit events for monitoring and integration:

api/core/workflow/entities/workflow_execution.py

@@ -5,6 +5,8 @@ Models are independent of the storage mechanism and don't contain
 implementation details like tenant_id, app_id, etc.
 """
 
+from __future__ import annotations
+
 from collections.abc import Mapping
 from datetime import datetime
 from typing import Any
@@ -59,7 +61,7 @@ class WorkflowExecution(BaseModel):
         graph: Mapping[str, Any],
         inputs: Mapping[str, Any],
         started_at: datetime,
-    ) -> "WorkflowExecution":
+    ) -> WorkflowExecution:
         return WorkflowExecution(
             id_=id_,
             workflow_id=workflow_id,

api/core/workflow/graph/graph.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 import logging
 from collections import defaultdict
 from collections.abc import Mapping, Sequence
@@ -175,7 +177,7 @@ class Graph:
     def _create_node_instances(
         cls,
         node_configs_map: dict[str, dict[str, object]],
-        node_factory: "NodeFactory",
+        node_factory: NodeFactory,
     ) -> dict[str, Node]:
         """
         Create node instances from configurations using the node factory.
@@ -197,7 +199,7 @@ class Graph:
         return nodes
 
     @classmethod
-    def new(cls) -> "GraphBuilder":
+    def new(cls) -> GraphBuilder:
         """Create a fluent builder for assembling a graph programmatically."""
 
         return GraphBuilder(graph_cls=cls)
@@ -284,9 +286,9 @@ class Graph:
         cls,
         *,
         graph_config: Mapping[str, object],
-        node_factory: "NodeFactory",
+        node_factory: NodeFactory,
         root_node_id: str | None = None,
-    ) -> "Graph":
+    ) -> Graph:
         """
         Initialize graph
 
@@ -383,7 +385,7 @@ class GraphBuilder:
         self._edges: list[Edge] = []
         self._edge_counter = 0
 
-    def add_root(self, node: Node) -> "GraphBuilder":
+    def add_root(self, node: Node) -> GraphBuilder:
         """Register the root node. Must be called exactly once."""
 
         if self._nodes:
@@ -398,7 +400,7 @@ class GraphBuilder:
         *,
         from_node_id: str | None = None,
         source_handle: str = "source",
-    ) -> "GraphBuilder":
+    ) -> GraphBuilder:
         """Append a node and connect it from the specified predecessor."""
 
         if not self._nodes:
@@ -419,7 +421,7 @@ class GraphBuilder:
 
         return self
 
-    def connect(self, *, tail: str, head: str, source_handle: str = "source") -> "GraphBuilder":
+    def connect(self, *, tail: str, head: str, source_handle: str = "source") -> GraphBuilder:
         """Connect two existing nodes without adding a new node."""
 
         if tail not in self._nodes_by_id:

api/core/workflow/graph_engine/command_channels/redis_channel.py

@@ -9,7 +9,7 @@ Each instance uses a unique key for its command queue.
 import json
 from typing import TYPE_CHECKING, Any, final
 
-from ..entities.commands import AbortCommand, CommandType, GraphEngineCommand, PauseCommand
+from ..entities.commands import AbortCommand, CommandType, GraphEngineCommand, PauseCommand, UpdateVariablesCommand
 
 if TYPE_CHECKING:
     from extensions.ext_redis import RedisClientWrapper
@@ -113,6 +113,8 @@ class RedisChannel:
                 return AbortCommand.model_validate(data)
             if command_type == CommandType.PAUSE:
                 return PauseCommand.model_validate(data)
+            if command_type == CommandType.UPDATE_VARIABLES:
+                return UpdateVariablesCommand.model_validate(data)
 
             # For other command types, use base class
             return GraphEngineCommand.model_validate(data)

api/core/workflow/graph_engine/command_processing/__init__.py

@@ -5,11 +5,12 @@ This package handles external commands sent to the engine
 during execution.
 """
 
-from .command_handlers import AbortCommandHandler, PauseCommandHandler
+from .command_handlers import AbortCommandHandler, PauseCommandHandler, UpdateVariablesCommandHandler
 from .command_processor import CommandProcessor
 
 __all__ = [
     "AbortCommandHandler",
     "CommandProcessor",
     "PauseCommandHandler",
+    "UpdateVariablesCommandHandler",
 ]

api/core/workflow/graph_engine/command_processing/command_handlers.py

@@ -4,9 +4,10 @@ from typing import final
 from typing_extensions import override
 
 from core.workflow.entities.pause_reason import SchedulingPause
+from core.workflow.runtime import VariablePool
 
 from ..domain.graph_execution import GraphExecution
-from ..entities.commands import AbortCommand, GraphEngineCommand, PauseCommand
+from ..entities.commands import AbortCommand, GraphEngineCommand, PauseCommand, UpdateVariablesCommand
 from .command_processor import CommandHandler
 
 logger = logging.getLogger(__name__)
@@ -31,3 +32,25 @@ class PauseCommandHandler(CommandHandler):
         reason = command.reason
         pause_reason = SchedulingPause(message=reason)
         execution.pause(pause_reason)
+
+
+@final
+class UpdateVariablesCommandHandler(CommandHandler):
+    def __init__(self, variable_pool: VariablePool) -> None:
+        self._variable_pool = variable_pool
+
+    @override
+    def handle(self, command: GraphEngineCommand, execution: GraphExecution) -> None:
+        assert isinstance(command, UpdateVariablesCommand)
+        for update in command.updates:
+            try:
+                variable = update.value
+                self._variable_pool.add(variable.selector, variable)
+                logger.debug("Updated variable %s for workflow %s", variable.selector, execution.workflow_id)
+            except ValueError as exc:
+                logger.warning(
+                    "Skipping invalid variable selector %s for workflow %s: %s",
+                    getattr(update.value, "selector", None),
+                    execution.workflow_id,
+                    exc,
+                )

api/core/workflow/graph_engine/entities/commands.py

@@ -5,17 +5,21 @@ This module defines command types that can be sent to a running GraphEngine
 instance to control its execution flow.
 """
 
-from enum import StrEnum
+from collections.abc import Sequence
+from enum import StrEnum, auto
 from typing import Any
 
 from pydantic import BaseModel, Field
 
+from core.variables.variables import VariableUnion
+
 
 class CommandType(StrEnum):
     """Types of commands that can be sent to GraphEngine."""
 
-    ABORT = "abort"
-    PAUSE = "pause"
+    ABORT = auto()
+    PAUSE = auto()
+    UPDATE_VARIABLES = auto()
 
 
 class GraphEngineCommand(BaseModel):
@@ -37,3 +41,16 @@ class PauseCommand(GraphEngineCommand):
 
     command_type: CommandType = Field(default=CommandType.PAUSE, description="Type of command")
     reason: str = Field(default="unknown reason", description="reason for pause")
+
+
+class VariableUpdate(BaseModel):
+    """Represents a single variable update instruction."""
+
+    value: VariableUnion = Field(description="New variable value")
+
+
+class UpdateVariablesCommand(GraphEngineCommand):
+    """Command to update a group of variables in the variable pool."""
+
+    command_type: CommandType = Field(default=CommandType.UPDATE_VARIABLES, description="Type of command")
+    updates: Sequence[VariableUpdate] = Field(default_factory=list, description="Variable updates")

api/core/workflow/graph_engine/graph_engine.py

@@ -5,9 +5,12 @@ This engine uses a modular architecture with separated packages following
 Domain-Driven Design principles for improved maintainability and testability.
 """
 
+from __future__ import annotations
+
 import contextvars
 import logging
 import queue
+import threading
 from collections.abc import Generator
 from typing import TYPE_CHECKING, cast, final
 
@@ -30,8 +33,13 @@ from core.workflow.runtime import GraphRuntimeState, ReadOnlyGraphRuntimeStateWr
 if TYPE_CHECKING:  # pragma: no cover - used only for static analysis
     from core.workflow.runtime.graph_runtime_state import GraphProtocol
 
-from .command_processing import AbortCommandHandler, CommandProcessor, PauseCommandHandler
-from .entities.commands import AbortCommand, PauseCommand
+from .command_processing import (
+    AbortCommandHandler,
+    CommandProcessor,
+    PauseCommandHandler,
+    UpdateVariablesCommandHandler,
+)
+from .entities.commands import AbortCommand, PauseCommand, UpdateVariablesCommand
 from .error_handler import ErrorHandler
 from .event_management import EventHandler, EventManager
 from .graph_state_manager import GraphStateManager
@@ -70,10 +78,13 @@ class GraphEngine:
         scale_down_idle_time: float | None = None,
     ) -> None:
         """Initialize the graph engine with all subsystems and dependencies."""
+        # stop event
+        self._stop_event = threading.Event()
 
         # Bind runtime state to current workflow context
         self._graph = graph
         self._graph_runtime_state = graph_runtime_state
+        self._graph_runtime_state.stop_event = self._stop_event
         self._graph_runtime_state.configure(graph=cast("GraphProtocol", graph))
         self._command_channel = command_channel
 
@@ -140,6 +151,9 @@ class GraphEngine:
         pause_handler = PauseCommandHandler()
         self._command_processor.register_handler(PauseCommand, pause_handler)
 
+        update_variables_handler = UpdateVariablesCommandHandler(self._graph_runtime_state.variable_pool)
+        self._command_processor.register_handler(UpdateVariablesCommand, update_variables_handler)
+
         # === Extensibility ===
         # Layers allow plugins to extend engine functionality
         self._layers: list[GraphEngineLayer] = []
@@ -169,6 +183,7 @@ class GraphEngine:
             max_workers=self._max_workers,
             scale_up_threshold=self._scale_up_threshold,
             scale_down_idle_time=self._scale_down_idle_time,
+            stop_event=self._stop_event,
         )
 
         # === Orchestration ===
@@ -199,6 +214,7 @@ class GraphEngine:
             event_handler=self._event_handler_registry,
             execution_coordinator=self._execution_coordinator,
             event_emitter=self._event_manager,
+            stop_event=self._stop_event,
         )
 
         # === Validation ===
@@ -212,9 +228,16 @@ class GraphEngine:
             if id(node.graph_runtime_state) != expected_state_id:
                 raise ValueError(f"GraphRuntimeState consistency violation: Node '{node.id}' has a different instance")
 
-    def layer(self, layer: GraphEngineLayer) -> "GraphEngine":
+    def _bind_layer_context(
+        self,
+        layer: GraphEngineLayer,
+    ) -> None:
+        layer.initialize(ReadOnlyGraphRuntimeStateWrapper(self._graph_runtime_state), self._command_channel)
+
+    def layer(self, layer: GraphEngineLayer) -> GraphEngine:
         """Add a layer for extending functionality."""
         self._layers.append(layer)
+        self._bind_layer_context(layer)
         return self
 
     def run(self) -> Generator[GraphEngineEvent, None, None]:
@@ -301,14 +324,7 @@ class GraphEngine:
     def _initialize_layers(self) -> None:
         """Initialize layers with context."""
         self._event_manager.set_layers(self._layers)
-        # Create a read-only wrapper for the runtime state
-        read_only_state = ReadOnlyGraphRuntimeStateWrapper(self._graph_runtime_state)
         for layer in self._layers:
-            try:
-                layer.initialize(read_only_state, self._command_channel)
-            except Exception as e:
-                logger.warning("Failed to initialize layer %s: %s", layer.__class__.__name__, e)
-
             try:
                 layer.on_graph_start()
             except Exception as e:
@@ -316,6 +332,7 @@ class GraphEngine:
 
     def _start_execution(self, *, resume: bool = False) -> None:
         """Start execution subsystems."""
+        self._stop_event.clear()
         paused_nodes: list[str] = []
         if resume:
             paused_nodes = self._graph_runtime_state.consume_paused_nodes()
@@ -343,13 +360,12 @@ class GraphEngine:
 
     def _stop_execution(self) -> None:
         """Stop execution subsystems."""
+        self._stop_event.set()
         self._dispatcher.stop()
         self._worker_pool.stop()
         # Don't mark complete here as the dispatcher already does it
 
         # Notify layers
-        logger = logging.getLogger(__name__)
-
         for layer in self._layers:
             try:
                 layer.on_graph_end(self._graph_execution.error)

api/core/workflow/graph_engine/layers/README.md

@@ -8,7 +8,7 @@ Pluggable middleware for engine extensions.
 
 Abstract base class for layers.
 
-- `initialize()` - Receive runtime context
+- `initialize()` - Receive runtime context (runtime state is bound here and always available to hooks)
 - `on_graph_start()` - Execution start hook
 - `on_event()` - Process all events
 - `on_graph_end()` - Execution end hook
@@ -34,6 +34,9 @@ engine.layer(debug_layer)
 engine.run()
 ```
 
+`engine.layer()` binds the read-only runtime state before execution, so
+`graph_runtime_state` is always available inside layer hooks.
+
 ## Custom Layers
 
 ```python

api/core/workflow/graph_engine/layers/base.py

@@ -13,6 +13,14 @@ from core.workflow.nodes.base.node import Node
 from core.workflow.runtime import ReadOnlyGraphRuntimeState
 
 
+class GraphEngineLayerNotInitializedError(Exception):
+    """Raised when a layer's runtime state is accessed before initialization."""
+
+    def __init__(self, layer_name: str | None = None) -> None:
+        name = layer_name or "GraphEngineLayer"
+        super().__init__(f"{name} runtime state is not initialized. Bind the layer to a GraphEngine before access.")
+
+
 class GraphEngineLayer(ABC):
     """
     Abstract base class for GraphEngine layers.
@@ -28,22 +36,27 @@ class GraphEngineLayer(ABC):
 
     def __init__(self) -> None:
         """Initialize the layer. Subclasses can override with custom parameters."""
-        self.graph_runtime_state: ReadOnlyGraphRuntimeState | None = None
+        self._graph_runtime_state: ReadOnlyGraphRuntimeState | None = None
         self.command_channel: CommandChannel | None = None
 
+    @property
+    def graph_runtime_state(self) -> ReadOnlyGraphRuntimeState:
+        if self._graph_runtime_state is None:
+            raise GraphEngineLayerNotInitializedError(type(self).__name__)
+        return self._graph_runtime_state
+
     def initialize(self, graph_runtime_state: ReadOnlyGraphRuntimeState, command_channel: CommandChannel) -> None:
         """
         Initialize the layer with engine dependencies.
 
-        Called by GraphEngine before execution starts to inject the read-only runtime state
-        and command channel. This allows layers to observe engine context and send
-        commands, but prevents direct state modification.
-
+        Called by GraphEngine to inject the read-only runtime state and command channel.
+        This is invoked when the layer is registered with a `GraphEngine` instance.
+        Implementations should be idempotent.
         Args:
             graph_runtime_state: Read-only view of the runtime state
             command_channel: Channel for sending commands to the engine
         """
-        self.graph_runtime_state = graph_runtime_state
+        self._graph_runtime_state = graph_runtime_state
         self.command_channel = command_channel
 
     @abstractmethod

api/core/workflow/graph_engine/layers/debug_logging.py

@@ -109,10 +109,8 @@ class DebugLoggingLayer(GraphEngineLayer):
         self.logger.info("=" * 80)
         self.logger.info("🚀 GRAPH EXECUTION STARTED")
         self.logger.info("=" * 80)
-
-        if self.graph_runtime_state:
-            # Log initial state
-            self.logger.info("Initial State:")
+        # Log initial state
+        self.logger.info("Initial State:")
 
     @override
     def on_event(self, event: GraphEngineEvent) -> None:
@@ -243,8 +241,7 @@ class DebugLoggingLayer(GraphEngineLayer):
         self.logger.info("  Node retries: %s", self.retry_count)
 
         # Log final state if available
-        if self.graph_runtime_state and self.include_outputs:
-            if self.graph_runtime_state.outputs:
-                self.logger.info("Final outputs: %s", self._format_dict(self.graph_runtime_state.outputs))
+        if self.include_outputs and self.graph_runtime_state.outputs:
+            self.logger.info("Final outputs: %s", self._format_dict(self.graph_runtime_state.outputs))
 
         self.logger.info("=" * 80)

api/core/workflow/graph_engine/layers/persistence.py

@@ -337,8 +337,6 @@ class WorkflowPersistenceLayer(GraphEngineLayer):
         if update_finished:
             execution.finished_at = naive_utc_now()
         runtime_state = self.graph_runtime_state
-        if runtime_state is None:
-            return
         execution.total_tokens = runtime_state.total_tokens
         execution.total_steps = runtime_state.node_run_steps
         execution.outputs = execution.outputs or runtime_state.outputs
@@ -404,6 +402,4 @@ class WorkflowPersistenceLayer(GraphEngineLayer):
 
     def _system_variables(self) -> Mapping[str, Any]:
         runtime_state = self.graph_runtime_state
-        if runtime_state is None:
-            return {}
         return runtime_state.variable_pool.get_by_prefix(SYSTEM_VARIABLE_NODE_ID)

api/core/workflow/graph_engine/manager.py

@@ -3,14 +3,20 @@ GraphEngine Manager for sending control commands via Redis channel.
 
 This module provides a simplified interface for controlling workflow executions
 using the new Redis command channel, without requiring user permission checks.
-Supports stop, pause, and resume operations.
 """
 
 import logging
+from collections.abc import Sequence
 from typing import final
 
 from core.workflow.graph_engine.command_channels.redis_channel import RedisChannel
-from core.workflow.graph_engine.entities.commands import AbortCommand, GraphEngineCommand, PauseCommand
+from core.workflow.graph_engine.entities.commands import (
+    AbortCommand,
+    GraphEngineCommand,
+    PauseCommand,
+    UpdateVariablesCommand,
+    VariableUpdate,
+)
 from extensions.ext_redis import redis_client
 
 logger = logging.getLogger(__name__)
@@ -23,7 +29,6 @@ class GraphEngineManager:
 
     This class provides a simple interface for controlling workflow executions
     by sending commands through Redis channels, without user validation.
-    Supports stop and pause operations.
     """
 
     @staticmethod
@@ -45,6 +50,16 @@ class GraphEngineManager:
         pause_command = PauseCommand(reason=reason or "User requested pause")
         GraphEngineManager._send_command(task_id, pause_command)
 
+    @staticmethod
+    def send_update_variables_command(task_id: str, updates: Sequence[VariableUpdate]) -> None:
+        """Send a command to update variables in a running workflow."""
+
+        if not updates:
+            return
+
+        update_command = UpdateVariablesCommand(updates=updates)
+        GraphEngineManager._send_command(task_id, update_command)
+
     @staticmethod
     def _send_command(task_id: str, command: GraphEngineCommand) -> None:
         """Send a command to the workflow-specific Redis channel."""

api/core/workflow/graph_engine/orchestration/dispatcher.py

@@ -44,6 +44,7 @@ class Dispatcher:
         event_queue: queue.Queue[GraphNodeEventBase],
         event_handler: "EventHandler",
         execution_coordinator: ExecutionCoordinator,
+        stop_event: threading.Event,
         event_emitter: EventManager | None = None,
     ) -> None:
         """
@@ -61,7 +62,7 @@ class Dispatcher:
         self._event_emitter = event_emitter
 
         self._thread: threading.Thread | None = None
-        self._stop_event = threading.Event()
+        self._stop_event = stop_event
         self._start_time: float | None = None
 
     def start(self) -> None:
@@ -69,16 +70,14 @@ class Dispatcher:
         if self._thread and self._thread.is_alive():
             return
 
-        self._stop_event.clear()
         self._start_time = time.time()
         self._thread = threading.Thread(target=self._dispatcher_loop, name="GraphDispatcher", daemon=True)
         self._thread.start()
 
     def stop(self) -> None:
         """Stop the dispatcher thread."""
-        self._stop_event.set()
         if self._thread and self._thread.is_alive():
-            self._thread.join(timeout=10.0)
+            self._thread.join(timeout=2.0)
 
     def _dispatcher_loop(self) -> None:
         """Main dispatcher loop."""

api/core/workflow/graph_engine/ready_queue/factory.py

@@ -2,6 +2,8 @@
 Factory for creating ReadyQueue instances from serialized state.
 """
 
+from __future__ import annotations
+
 from typing import TYPE_CHECKING
 
 from .in_memory import InMemoryReadyQueue
@@ -11,7 +13,7 @@ if TYPE_CHECKING:
     from .protocol import ReadyQueue
 
 
-def create_ready_queue_from_state(state: ReadyQueueState) -> "ReadyQueue":
+def create_ready_queue_from_state(state: ReadyQueueState) -> ReadyQueue:
     """
     Create a ReadyQueue instance from a serialized state.
 

api/core/workflow/graph_engine/response_coordinator/session.py

@@ -5,6 +5,8 @@ This module contains the private ResponseSession class used internally
 by ResponseStreamCoordinator to manage streaming sessions.
 """
 
+from __future__ import annotations
+
 from dataclasses import dataclass
 
 from core.workflow.nodes.answer.answer_node import AnswerNode
@@ -27,7 +29,7 @@ class ResponseSession:
     index: int = 0  # Current position in the template segments
 
     @classmethod
-    def from_node(cls, node: Node) -> "ResponseSession":
+    def from_node(cls, node: Node) -> ResponseSession:
         """
         Create a ResponseSession from an AnswerNode or EndNode.
 

api/core/workflow/graph_engine/worker.py

@@ -42,6 +42,7 @@ class Worker(threading.Thread):
         event_queue: queue.Queue[GraphNodeEventBase],
         graph: Graph,
         layers: Sequence[GraphEngineLayer],
+        stop_event: threading.Event,
         worker_id: int = 0,
         flask_app: Flask | None = None,
         context_vars: contextvars.Context | None = None,
@@ -65,13 +66,16 @@ class Worker(threading.Thread):
         self._worker_id = worker_id
         self._flask_app = flask_app
         self._context_vars = context_vars
-        self._stop_event = threading.Event()
         self._last_task_time = time.time()
+        self._stop_event = stop_event
         self._layers = layers if layers is not None else []
 
     def stop(self) -> None:
-        """Signal the worker to stop processing."""
-        self._stop_event.set()
+        """Worker is controlled via shared stop_event from GraphEngine.
+
+        This method is a no-op retained for backward compatibility.
+        """
+        pass
 
     @property
     def is_idle(self) -> bool:

api/core/workflow/graph_engine/worker_management/worker_pool.py

@@ -41,6 +41,7 @@ class WorkerPool:
         event_queue: queue.Queue[GraphNodeEventBase],
         graph: Graph,
         layers: list[GraphEngineLayer],
+        stop_event: threading.Event,
         flask_app: "Flask | None" = None,
         context_vars: "Context | None" = None,
         min_workers: int | None = None,
@@ -81,6 +82,7 @@ class WorkerPool:
         self._worker_counter = 0
         self._lock = threading.RLock()
         self._running = False
+        self._stop_event = stop_event
 
         # No longer tracking worker states with callbacks to avoid lock contention
 
@@ -135,7 +137,7 @@ class WorkerPool:
             # Wait for workers to finish
             for worker in self._workers:
                 if worker.is_alive():
-                    worker.join(timeout=10.0)
+                    worker.join(timeout=2.0)
 
             self._workers.clear()
 
@@ -152,6 +154,7 @@ class WorkerPool:
             worker_id=worker_id,
             flask_app=self._flask_app,
             context_vars=self._context_vars,
+            stop_event=self._stop_event,
         )
 
         worker.start()

api/core/workflow/nodes/agent/agent_node.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 import json
 from collections.abc import Generator, Mapping, Sequence
 from typing import TYPE_CHECKING, Any, cast
@@ -167,7 +169,7 @@ class AgentNode(Node[AgentNodeData]):
         variable_pool: VariablePool,
         node_data: AgentNodeData,
         for_log: bool = False,
-        strategy: "PluginAgentStrategy",
+        strategy: PluginAgentStrategy,
     ) -> dict[str, Any]:
         """
         Generate parameters based on the given tool parameters, variable pool, and node data.
@@ -328,7 +330,7 @@ class AgentNode(Node[AgentNodeData]):
     def _generate_credentials(
         self,
         parameters: dict[str, Any],
-    ) -> "InvokeCredentials":
+    ) -> InvokeCredentials:
         """
         Generate credentials based on the given agent parameters.
         """
@@ -442,9 +444,7 @@ class AgentNode(Node[AgentNodeData]):
                     model_schema.features.remove(feature)
         return model_schema
 
-    def _filter_mcp_type_tool(
-        self, strategy: "PluginAgentStrategy", tools: list[dict[str, Any]]
-    ) -> list[dict[str, Any]]:
+    def _filter_mcp_type_tool(self, strategy: PluginAgentStrategy, tools: list[dict[str, Any]]) -> list[dict[str, Any]]:
         """
         Filter MCP type tool
         :param strategy: plugin agent strategy

api/core/workflow/nodes/base/entities.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 import json
 from abc import ABC
 from builtins import type as type_
@@ -111,7 +113,7 @@ class DefaultValue(BaseModel):
             raise DefaultValueTypeError(f"Cannot convert to number: {value}")
 
     @model_validator(mode="after")
-    def validate_value_type(self) -> "DefaultValue":
+    def validate_value_type(self) -> DefaultValue:
         # Type validation configuration
         type_validators = {
             DefaultValueType.STRING: {

api/core/workflow/nodes/base/node.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 import importlib
 import logging
 import operator
@@ -59,7 +61,7 @@ logger = logging.getLogger(__name__)
 
 
 class Node(Generic[NodeDataT]):
-    node_type: ClassVar["NodeType"]
+    node_type: ClassVar[NodeType]
     execution_type: NodeExecutionType = NodeExecutionType.EXECUTABLE
     _node_data_type: ClassVar[type[BaseNodeData]] = BaseNodeData
 
@@ -198,14 +200,14 @@ class Node(Generic[NodeDataT]):
         return None
 
     # Global registry populated via __init_subclass__
-    _registry: ClassVar[dict["NodeType", dict[str, type["Node"]]]] = {}
+    _registry: ClassVar[dict[NodeType, dict[str, type[Node]]]] = {}
 
     def __init__(
         self,
         id: str,
         config: Mapping[str, Any],
-        graph_init_params: "GraphInitParams",
-        graph_runtime_state: "GraphRuntimeState",
+        graph_init_params: GraphInitParams,
+        graph_runtime_state: GraphRuntimeState,
     ) -> None:
         self._graph_init_params = graph_init_params
         self.id = id
@@ -241,7 +243,7 @@ class Node(Generic[NodeDataT]):
         return
 
     @property
-    def graph_init_params(self) -> "GraphInitParams":
+    def graph_init_params(self) -> GraphInitParams:
         return self._graph_init_params
 
     @property
@@ -264,6 +266,10 @@ class Node(Generic[NodeDataT]):
         """
         raise NotImplementedError
 
+    def _should_stop(self) -> bool:
+        """Check if execution should be stopped."""
+        return self.graph_runtime_state.stop_event.is_set()
+
     def run(self) -> Generator[GraphNodeEventBase, None, None]:
         execution_id = self.ensure_execution_id()
         self._start_at = naive_utc_now()
@@ -332,6 +338,21 @@ class Node(Generic[NodeDataT]):
                     yield event
                 else:
                     yield event
+
+                if self._should_stop():
+                    error_message = "Execution cancelled"
+                    yield NodeRunFailedEvent(
+                        id=self.execution_id,
+                        node_id=self._node_id,
+                        node_type=self.node_type,
+                        start_at=self._start_at,
+                        node_run_result=NodeRunResult(
+                            status=WorkflowNodeExecutionStatus.FAILED,
+                            error=error_message,
+                        ),
+                        error=error_message,
+                    )
+                    return
         except Exception as e:
             logger.exception("Node %s failed to run", self._node_id)
             result = NodeRunResult(
@@ -438,7 +459,7 @@ class Node(Generic[NodeDataT]):
         raise NotImplementedError("subclasses of BaseNode must implement `version` method.")
 
     @classmethod
-    def get_node_type_classes_mapping(cls) -> Mapping["NodeType", Mapping[str, type["Node"]]]:
+    def get_node_type_classes_mapping(cls) -> Mapping[NodeType, Mapping[str, type[Node]]]:
         """Return mapping of NodeType -> {version -> Node subclass} using __init_subclass__ registry.
 
         Import all modules under core.workflow.nodes so subclasses register themselves on import.

api/core/workflow/nodes/base/template.py

@@ -4,6 +4,8 @@ This module provides a unified template structure for both Answer and End nodes,
 similar to SegmentGroup but focused on template representation without values.
 """
 
+from __future__ import annotations
+
 from abc import ABC, abstractmethod
 from collections.abc import Sequence
 from dataclasses import dataclass
@@ -58,7 +60,7 @@ class Template:
     segments: list[TemplateSegmentUnion]
 
     @classmethod
-    def from_answer_template(cls, template_str: str) -> "Template":
+    def from_answer_template(cls, template_str: str) -> Template:
         """Create a Template from an Answer node template string.
 
         Example:
@@ -107,7 +109,7 @@ class Template:
         return cls(segments=segments)
 
     @classmethod
-    def from_end_outputs(cls, outputs_config: list[dict[str, Any]]) -> "Template":
+    def from_end_outputs(cls, outputs_config: list[dict[str, Any]]) -> Template:
         """Create a Template from an End node outputs configuration.
 
         End nodes are treated as templates of concatenated variables with newlines.

api/core/workflow/nodes/llm/node.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 import base64
 import io
 import json
@@ -113,7 +115,7 @@ class LLMNode(Node[LLMNodeData]):
 
     # Instance attributes specific to LLMNode.
     # Output variable for file
-    _file_outputs: list["File"]
+    _file_outputs: list[File]
 
     _llm_file_saver: LLMFileSaver
 
@@ -121,8 +123,8 @@ class LLMNode(Node[LLMNodeData]):
         self,
         id: str,
         config: Mapping[str, Any],
-        graph_init_params: "GraphInitParams",
-        graph_runtime_state: "GraphRuntimeState",
+        graph_init_params: GraphInitParams,
+        graph_runtime_state: GraphRuntimeState,
         *,
         llm_file_saver: LLMFileSaver | None = None,
     ):
@@ -361,7 +363,7 @@ class LLMNode(Node[LLMNodeData]):
         structured_output_enabled: bool,
         structured_output: Mapping[str, Any] | None = None,
         file_saver: LLMFileSaver,
-        file_outputs: list["File"],
+        file_outputs: list[File],
         node_id: str,
         node_type: NodeType,
         reasoning_format: Literal["separated", "tagged"] = "tagged",
@@ -415,7 +417,7 @@ class LLMNode(Node[LLMNodeData]):
         *,
         invoke_result: LLMResult | Generator[LLMResultChunk | LLMStructuredOutput, None, None],
         file_saver: LLMFileSaver,
-        file_outputs: list["File"],
+        file_outputs: list[File],
         node_id: str,
         node_type: NodeType,
         reasoning_format: Literal["separated", "tagged"] = "tagged",
@@ -525,7 +527,7 @@ class LLMNode(Node[LLMNodeData]):
         )
 
     @staticmethod
-    def _image_file_to_markdown(file: "File", /):
+    def _image_file_to_markdown(file: File, /):
         text_chunk = f"![]({file.generate_url()})"
         return text_chunk
 
@@ -774,7 +776,7 @@ class LLMNode(Node[LLMNodeData]):
     def fetch_prompt_messages(
         *,
         sys_query: str | None = None,
-        sys_files: Sequence["File"],
+        sys_files: Sequence[File],
         context: str | None = None,
         memory: TokenBufferMemory | None = None,
         model_config: ModelConfigWithCredentialsEntity,
@@ -785,7 +787,7 @@ class LLMNode(Node[LLMNodeData]):
         variable_pool: VariablePool,
         jinja2_variables: Sequence[VariableSelector],
         tenant_id: str,
-        context_files: list["File"] | None = None,
+        context_files: list[File] | None = None,
     ) -> tuple[Sequence[PromptMessage], Sequence[str] | None]:
         prompt_messages: list[PromptMessage] = []
 
@@ -1137,7 +1139,7 @@ class LLMNode(Node[LLMNodeData]):
         *,
         invoke_result: LLMResult | LLMResultWithStructuredOutput,
         saver: LLMFileSaver,
-        file_outputs: list["File"],
+        file_outputs: list[File],
         reasoning_format: Literal["separated", "tagged"] = "tagged",
         request_latency: float | None = None,
     ) -> ModelInvokeCompletedEvent:
@@ -1179,7 +1181,7 @@ class LLMNode(Node[LLMNodeData]):
         *,
         content: ImagePromptMessageContent,
         file_saver: LLMFileSaver,
-    ) -> "File":
+    ) -> File:
         """_save_multimodal_output saves multi-modal contents generated by LLM plugins.
 
         There are two kinds of multimodal outputs:
@@ -1229,7 +1231,7 @@ class LLMNode(Node[LLMNodeData]):
         *,
         contents: str | list[PromptMessageContentUnionTypes] | None,
         file_saver: LLMFileSaver,
-        file_outputs: list["File"],
+        file_outputs: list[File],
     ) -> Generator[str, None, None]:
         """Convert intermediate prompt messages into strings and yield them to the caller.
 

api/core/workflow/nodes/node_factory.py

@@ -113,7 +113,6 @@ class DifyNodeFactory(NodeFactory):
                 code_providers=self._code_providers,
                 code_limits=self._code_limits,
             )
-
         if node_type == NodeType.TEMPLATE_TRANSFORM:
             return TemplateTransformNode(
                 id=node_id,

api/core/workflow/nodes/variable_assigner/common/impl.py

@@ -1,28 +0,0 @@
-from sqlalchemy import select
-from sqlalchemy.orm import Session
-
-from core.variables.variables import Variable
-from extensions.ext_database import db
-from models import ConversationVariable
-
-from .exc import VariableOperatorNodeError
-
-
-class ConversationVariableUpdaterImpl:
-    def update(self, conversation_id: str, variable: Variable):
-        stmt = select(ConversationVariable).where(
-            ConversationVariable.id == variable.id, ConversationVariable.conversation_id == conversation_id
-        )
-        with Session(db.engine) as session:
-            row = session.scalar(stmt)
-            if not row:
-                raise VariableOperatorNodeError("conversation variable not found in the database")
-            row.data = variable.model_dump_json()
-            session.commit()
-
-    def flush(self):
-        pass
-
-
-def conversation_variable_updater_factory() -> ConversationVariableUpdaterImpl:
-    return ConversationVariableUpdaterImpl()

api/core/workflow/nodes/variable_assigner/v1/node.py

@@ -1,9 +1,8 @@
-from collections.abc import Callable, Mapping, Sequence
-from typing import TYPE_CHECKING, Any, TypeAlias
+from collections.abc import Mapping, Sequence
+from typing import TYPE_CHECKING, Any
 
 from core.variables import SegmentType, Variable
 from core.workflow.constants import CONVERSATION_VARIABLE_NODE_ID
-from core.workflow.conversation_variable_updater import ConversationVariableUpdater
 from core.workflow.entities import GraphInitParams
 from core.workflow.enums import NodeType, WorkflowNodeExecutionStatus
 from core.workflow.node_events import NodeRunResult
@@ -11,19 +10,14 @@ from core.workflow.nodes.base.node import Node
 from core.workflow.nodes.variable_assigner.common import helpers as common_helpers
 from core.workflow.nodes.variable_assigner.common.exc import VariableOperatorNodeError
 
-from ..common.impl import conversation_variable_updater_factory
 from .node_data import VariableAssignerData, WriteMode
 
 if TYPE_CHECKING:
     from core.workflow.runtime import GraphRuntimeState
 
 
-_CONV_VAR_UPDATER_FACTORY: TypeAlias = Callable[[], ConversationVariableUpdater]
-
-
 class VariableAssignerNode(Node[VariableAssignerData]):
     node_type = NodeType.VARIABLE_ASSIGNER
-    _conv_var_updater_factory: _CONV_VAR_UPDATER_FACTORY
 
     def __init__(
         self,
@@ -31,7 +25,6 @@ class VariableAssignerNode(Node[VariableAssignerData]):
         config: Mapping[str, Any],
         graph_init_params: "GraphInitParams",
         graph_runtime_state: "GraphRuntimeState",
-        conv_var_updater_factory: _CONV_VAR_UPDATER_FACTORY = conversation_variable_updater_factory,
     ):
         super().__init__(
             id=id,
@@ -39,7 +32,6 @@ class VariableAssignerNode(Node[VariableAssignerData]):
             graph_init_params=graph_init_params,
             graph_runtime_state=graph_runtime_state,
         )
-        self._conv_var_updater_factory = conv_var_updater_factory
 
     @classmethod
     def version(cls) -> str:
@@ -96,16 +88,7 @@ class VariableAssignerNode(Node[VariableAssignerData]):
         # Over write the variable.
         self.graph_runtime_state.variable_pool.add(assigned_variable_selector, updated_variable)
 
-        # TODO: Move database operation to the pipeline.
-        # Update conversation variable.
-        conversation_id = self.graph_runtime_state.variable_pool.get(["sys", "conversation_id"])
-        if not conversation_id:
-            raise VariableOperatorNodeError("conversation_id not found")
-        conv_var_updater = self._conv_var_updater_factory()
-        conv_var_updater.update(conversation_id=conversation_id.text, variable=updated_variable)
-        conv_var_updater.flush()
         updated_variables = [common_helpers.variable_to_processed_data(assigned_variable_selector, updated_variable)]
-
         return NodeRunResult(
             status=WorkflowNodeExecutionStatus.SUCCEEDED,
             inputs={

api/core/workflow/nodes/variable_assigner/v2/node.py

@@ -1,24 +1,20 @@
 import json
 from collections.abc import Mapping, MutableMapping, Sequence
-from typing import Any, cast
+from typing import TYPE_CHECKING, Any
 
-from core.app.entities.app_invoke_entities import InvokeFrom
 from core.variables import SegmentType, Variable
 from core.variables.consts import SELECTORS_LENGTH
 from core.workflow.constants import CONVERSATION_VARIABLE_NODE_ID
-from core.workflow.conversation_variable_updater import ConversationVariableUpdater
 from core.workflow.enums import NodeType, WorkflowNodeExecutionStatus
 from core.workflow.node_events import NodeRunResult
 from core.workflow.nodes.base.node import Node
 from core.workflow.nodes.variable_assigner.common import helpers as common_helpers
 from core.workflow.nodes.variable_assigner.common.exc import VariableOperatorNodeError
-from core.workflow.nodes.variable_assigner.common.impl import conversation_variable_updater_factory
 
 from . import helpers
 from .entities import VariableAssignerNodeData, VariableOperationItem
 from .enums import InputType, Operation
 from .exc import (
-    ConversationIDNotFoundError,
     InputTypeNotSupportedError,
     InvalidDataError,
     InvalidInputValueError,
@@ -26,6 +22,10 @@ from .exc import (
     VariableNotFoundError,
 )
 
+if TYPE_CHECKING:
+    from core.workflow.entities import GraphInitParams
+    from core.workflow.runtime import GraphRuntimeState
+
 
 def _target_mapping_from_item(mapping: MutableMapping[str, Sequence[str]], node_id: str, item: VariableOperationItem):
     selector_node_id = item.variable_selector[0]
@@ -53,6 +53,20 @@ def _source_mapping_from_item(mapping: MutableMapping[str, Sequence[str]], node_
 class VariableAssignerNode(Node[VariableAssignerNodeData]):
     node_type = NodeType.VARIABLE_ASSIGNER
 
+    def __init__(
+        self,
+        id: str,
+        config: Mapping[str, Any],
+        graph_init_params: "GraphInitParams",
+        graph_runtime_state: "GraphRuntimeState",
+    ):
+        super().__init__(
+            id=id,
+            config=config,
+            graph_init_params=graph_init_params,
+            graph_runtime_state=graph_runtime_state,
+        )
+
     def blocks_variable_output(self, variable_selectors: set[tuple[str, ...]]) -> bool:
         """
         Check if this Variable Assigner node blocks the output of specific variables.
@@ -70,9 +84,6 @@ class VariableAssignerNode(Node[VariableAssignerNodeData]):
 
         return False
 
-    def _conv_var_updater_factory(self) -> ConversationVariableUpdater:
-        return conversation_variable_updater_factory()
-
     @classmethod
     def version(cls) -> str:
         return "2"
@@ -179,26 +190,12 @@ class VariableAssignerNode(Node[VariableAssignerNodeData]):
         # remove the duplicated items first.
         updated_variable_selectors = list(set(map(tuple, updated_variable_selectors)))
 
-        conv_var_updater = self._conv_var_updater_factory()
-        # Update variables
         for selector in updated_variable_selectors:
             variable = self.graph_runtime_state.variable_pool.get(selector)
             if not isinstance(variable, Variable):
                 raise VariableNotFoundError(variable_selector=selector)
             process_data[variable.name] = variable.value
 
-            if variable.selector[0] == CONVERSATION_VARIABLE_NODE_ID:
-                conversation_id = self.graph_runtime_state.variable_pool.get(["sys", "conversation_id"])
-                if not conversation_id:
-                    if self.invoke_from != InvokeFrom.DEBUGGER:
-                        raise ConversationIDNotFoundError
-                else:
-                    conversation_id = conversation_id.value
-                    conv_var_updater.update(
-                        conversation_id=cast(str, conversation_id),
-                        variable=variable,
-                    )
-        conv_var_updater.flush()
         updated_variables = [
             common_helpers.variable_to_processed_data(selector, seg)
             for selector in updated_variable_selectors

api/core/workflow/repositories/draft_variable_repository.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 import abc
 from collections.abc import Mapping
 from typing import Any, Protocol
@@ -23,7 +25,7 @@ class DraftVariableSaverFactory(Protocol):
         node_type: NodeType,
         node_execution_id: str,
         enclosing_node_id: str | None = None,
-    ) -> "DraftVariableSaver":
+    ) -> DraftVariableSaver:
         pass
 
 

api/core/workflow/runtime/graph_runtime_state.py

@@ -2,6 +2,7 @@ from __future__ import annotations
 
 import importlib
 import json
+import threading
 from collections.abc import Mapping, Sequence
 from copy import deepcopy
 from dataclasses import dataclass
@@ -168,6 +169,7 @@ class GraphRuntimeState:
         self._pending_response_coordinator_dump: str | None = None
         self._pending_graph_execution_workflow_id: str | None = None
         self._paused_nodes: set[str] = set()
+        self.stop_event: threading.Event = threading.Event()
 
         if graph is not None:
             self.attach_graph(graph)

api/core/workflow/runtime/graph_runtime_state_protocol.py

@@ -1,4 +1,4 @@
-from collections.abc import Mapping
+from collections.abc import Mapping, Sequence
 from typing import Any, Protocol
 
 from core.model_runtime.entities.llm_entities import LLMUsage
@@ -9,7 +9,7 @@ from core.workflow.system_variable import SystemVariableReadOnlyView
 class ReadOnlyVariablePool(Protocol):
     """Read-only interface for VariablePool."""
 
-    def get(self, node_id: str, variable_key: str) -> Segment | None:
+    def get(self, selector: Sequence[str], /) -> Segment | None:
         """Get a variable value (read-only)."""
         ...
 

api/core/workflow/runtime/read_only_wrappers.py

@@ -1,6 +1,6 @@
 from __future__ import annotations
 
-from collections.abc import Mapping
+from collections.abc import Mapping, Sequence
 from copy import deepcopy
 from typing import Any
 
@@ -18,9 +18,9 @@ class ReadOnlyVariablePoolWrapper:
     def __init__(self, variable_pool: VariablePool) -> None:
         self._variable_pool = variable_pool
 
-    def get(self, node_id: str, variable_key: str) -> Segment | None:
+    def get(self, selector: Sequence[str], /) -> Segment | None:
         """Return a copy of a variable value if present."""
-        value = self._variable_pool.get([node_id, variable_key])
+        value = self._variable_pool.get(selector)
         return deepcopy(value) if value is not None else None
 
     def get_all_by_node(self, node_id: str) -> Mapping[str, object]:

api/core/workflow/runtime/variable_pool.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 import re
 from collections import defaultdict
 from collections.abc import Mapping, Sequence
@@ -267,6 +269,6 @@ class VariablePool(BaseModel):
             self.add(selector, value)
 
     @classmethod
-    def empty(cls) -> "VariablePool":
+    def empty(cls) -> VariablePool:
         """Create an empty variable pool."""
         return cls(system_variables=SystemVariable.empty())

api/core/workflow/system_variable.py

@@ -1,3 +1,5 @@
+from __future__ import annotations
+
 from collections.abc import Mapping, Sequence
 from types import MappingProxyType
 from typing import Any
@@ -70,7 +72,7 @@ class SystemVariable(BaseModel):
         return data
 
     @classmethod
-    def empty(cls) -> "SystemVariable":
+    def empty(cls) -> SystemVariable:
         return cls()
 
     def to_dict(self) -> dict[SystemVariableKey, Any]:
@@ -114,7 +116,7 @@ class SystemVariable(BaseModel):
             d[SystemVariableKey.TIMESTAMP] = self.timestamp
         return d
 
-    def as_view(self) -> "SystemVariableReadOnlyView":
+    def as_view(self) -> SystemVariableReadOnlyView:
         return SystemVariableReadOnlyView(self)
 
 

api/docker/entrypoint.sh

@@ -3,8 +3,9 @@
 set -e
 
 # Set UTF-8 encoding to address potential encoding issues in containerized environments
-export LANG=${LANG:-en_US.UTF-8}
-export LC_ALL=${LC_ALL:-en_US.UTF-8}
+# Use C.UTF-8 which is universally available in all containers
+export LANG=${LANG:-C.UTF-8}
+export LC_ALL=${LC_ALL:-C.UTF-8}
 export PYTHONIOENCODING=${PYTHONIOENCODING:-utf-8}
 
 if [[ "${MIGRATION_ENABLED}" == "true" ]]; then