Use GPT5-Codex for PR Reviews (#566)

Signed-off-by: Glenn Jocher <[email protected]>
Co-authored-by: UltralyticsAssistant <[email protected]>

README.md

@@ -26,8 +26,8 @@ Ultralytics Actions automatically applies formats, updates, and enhancements usi
 - **Spell Check:** Common misspellings are caught using [codespell](https://github.com/codespell-project/codespell).
 - **Broken Links Check:** Broken links in documentation and Markdown files are identified using [Lychee](https://github.com/lycheeverse/lychee).
 - **PR Summary:** Concise Pull Request summaries are generated using [OpenAI](https://openai.com/) GPT-5, improving clarity and review efficiency.
-- **PR Review:** AI-powered inline code reviews identify critical bugs, security issues, and code quality concerns with suggested fixes.
-- **Auto-labeling:** Relevant labels are applied to issues and pull requests via [OpenAI](https://openai.com/) GPT-5 for intelligent categorization.
+- **PR Review:** AI-powered code reviews identify critical bugs, security issues, and code quality concerns with suggested fixes.
+- **Auto-labeling:** Applies relevant labels to issues and PRs via [OpenAI](https://openai.com/) GPT-5 for intelligent categorization.
 
 ## 🛠️ How It Works
 

actions/__init__.py

@@ -23,4 +23,4 @@
 #     ├── test_summarize_pr.py
 #     └── ...
 
-__version__ = "0.1.3"
+__version__ = "0.1.4"

actions/first_interaction.py

@@ -196,7 +196,7 @@ def main(*args, **kwargs):
 
         if summary := response.get("summary"):
             print("Updating PR description with summary...")
-            event.update_pr_description(number, SUMMARY_START + summary + "\n\n" + body)
+            event.update_pr_description(number, SUMMARY_START + summary)
         else:
             summary = body
 

actions/review_pr.py

@@ -125,7 +125,7 @@ def generate_pr_review(repository: str, diff_text: str, pr_title: str, pr_descri
     ]
 
     try:
-        response = get_completion(messages, reasoning_effort="medium")
+        response = get_completion(messages, reasoning_effort="medium", model="gpt-5-codex")
         print("\n" + "=" * 80 + f"\nFULL AI RESPONSE:\n{response}\n" + "=" * 80 + "\n")
 
         json_str = re.search(r"```(?:json)?\s*(\{.*?\})\s*```", response, re.DOTALL)

actions/utils/github_utils.py

@@ -281,10 +281,10 @@ class Action:
         start = "## 🛠️ PR Summary"
         if start in description:
             print("Existing PR Summary found, replacing.")
-            updated_description = description.split(start)[0] + new_summary
+            updated_description = description.split(start)[0].rstrip() + "\n\n" + new_summary
         else:
             print("PR Summary not found, appending.")
-            updated_description = description + "\n\n" + new_summary
+            updated_description = (description.rstrip() + "\n\n" + new_summary) if description.strip() else new_summary
 
         self.patch(url, json={"body": updated_description})
         self._pr_summary_cache = new_summary

actions/utils/openai_utils.py

@@ -112,6 +112,7 @@ def get_completion(
     temperature: float = 1.0,
     reasoning_effort: str = None,
     response_format: dict = None,
+    model: str = OPENAI_MODEL,
 ) -> str | dict:
     """Generates a completion using OpenAI's Responses API based on input messages."""
     assert OPENAI_API_KEY, "OpenAI API key is required."
@@ -122,8 +123,8 @@ def get_completion(
 
     max_retries = 2
     for attempt in range(max_retries + 2):
-        data = {"model": OPENAI_MODEL, "input": messages, "store": False, "temperature": temperature}
-        if "gpt-5" in OPENAI_MODEL:
+        data = {"model": model, "input": messages, "store": False, "temperature": temperature}
+        if "gpt-5" in model:
             data["reasoning"] = {"effort": reasoning_effort or "low"}
             # GPT-5 Responses API handles JSON via prompting, not format parameter