diff --git a/pkg/detectors/xai/xai.go b/pkg/detectors/xai/xai.go new file mode 100644 index 000000000..66512d0b4 --- /dev/null +++ b/pkg/detectors/xai/xai.go @@ -0,0 +1,124 @@ +package xai + +import ( + "context" + "encoding/json" + "fmt" + "io" + "net/http" + "strings" + + regexp "github.com/wasilibs/go-re2" + + "github.com/trufflesecurity/trufflehog/v3/pkg/common" + "github.com/trufflesecurity/trufflehog/v3/pkg/detectors" + "github.com/trufflesecurity/trufflehog/v3/pkg/pb/detectorspb" +) + +type Scanner struct { + client *http.Client +} + +// Ensure the Scanner satisfies the interface at compile time. +var _ detectors.Detector = (*Scanner)(nil) + +var ( + defaultClient = common.SaneHttpClient() + // Make sure that your group is surrounded in boundary characters such as below to reduce false positives. + keyPat = regexp.MustCompile(`\b(xai-[0-9a-zA-Z_]{80})\b`) +) + +// Keywords are used for efficiently pre-filtering chunks. +// Use identifiers in the secret preferably, or the provider name. +func (s Scanner) Keywords() []string { + return []string{"xai-"} +} + +// FromData will find and optionally verify Xai secrets in a given set of bytes. +func (s Scanner) FromData(ctx context.Context, verify bool, data []byte) (results []detectors.Result, err error) { + dataStr := string(data) + + keyMatches := make(map[string]struct{}) + for _, match := range keyPat.FindAllStringSubmatch(dataStr, -1) { + keyMatches[match[1]] = struct{}{} + } + + for match := range keyMatches { + s1 := detectors.Result{ + DetectorType: detectorspb.DetectorType_XAI, + Raw: []byte(match), + } + + if verify { + client := s.client + if client == nil { + client = defaultClient + } + + isVerified, extraData, verificationErr := verifyMatch(ctx, client, match) + s1.Verified = isVerified + s1.ExtraData = extraData + s1.SetVerificationError(verificationErr, match) + } + + results = append(results, s1) + } + + return +} + +func verifyMatch(ctx context.Context, client *http.Client, apiKey string) (bool, map[string]string, error) { + req, err := http.NewRequestWithContext(ctx, http.MethodGet, "https://api.x.ai/v1/api-key", nil) + if err != nil { + return false, nil, nil + } + + req.Header.Add("Content-Type", "application/json") + req.Header.Add("Authorization", "Bearer "+apiKey) + + res, err := client.Do(req) + if err != nil { + return false, nil, err + } + defer func() { + _, _ = io.Copy(io.Discard, res.Body) + _ = res.Body.Close() + }() + + switch res.StatusCode { + case http.StatusOK: + // Parse the API response for useful information like name and ACLs + var data struct { + Name string `json:"name"` + Acls []string `json:"acls"` + } + if err := json.NewDecoder(res.Body).Decode(&data); err != nil { + // The API Key is still verified, but there are parsing errors. + // Hence, return true for verified along with error. + return true, nil, fmt.Errorf("failed to decode response: %w", err) + } + + aclsStr := strings.Join(data.Acls, ",") + + // Convert the relevant fields into a map + result := map[string]string{ + "name": data.Name, + "acls": aclsStr, + } + + return true, result, nil + case http.StatusBadRequest, http.StatusUnauthorized: + // The secret is determinately not verified (nothing to do) + return false, nil, nil + default: + return false, nil, fmt.Errorf("unexpected HTTP response status %d", res.StatusCode) + } +} + +func (s Scanner) Type() detectorspb.DetectorType { + return detectorspb.DetectorType_XAI +} + +func (s Scanner) Description() string { + return "xAI is an AI company with the mission of advancing scientific discovery and gaining a deeper understanding of our universe." +} diff --git a/pkg/detectors/xai/xai_integration_test.go b/pkg/detectors/xai/xai_integration_test.go new file mode 100644 index 000000000..5bf5309f8 --- /dev/null +++ b/pkg/detectors/xai/xai_integration_test.go @@ -0,0 +1,161 @@ +//go:build detectors +// +build detectors + +package xai + +import ( + "context" + "fmt" + "testing" + "time" + + "github.com/google/go-cmp/cmp" + "github.com/google/go-cmp/cmp/cmpopts" + + "github.com/trufflesecurity/trufflehog/v3/pkg/common" + "github.com/trufflesecurity/trufflehog/v3/pkg/detectors" + "github.com/trufflesecurity/trufflehog/v3/pkg/pb/detectorspb" +) + +func TestXai_FromChunk(t *testing.T) { + ctx, cancel := context.WithTimeout(context.Background(), time.Second*5) + defer cancel() + testSecrets, err := common.GetSecret(ctx, "trufflehog-testing", "detectors5") + if err != nil { + t.Fatalf("could not get test secrets from GCP: %s", err) + } + secret := testSecrets.MustGetField("XAI") + inactiveSecret := testSecrets.MustGetField("XAI_INACTIVE") + + type args struct { + ctx context.Context + data []byte + verify bool + } + tests := []struct { + name string + s Scanner + args args + want []detectors.Result + wantErr bool + wantVerificationErr bool + }{ + { + name: "found, verified", + s: Scanner{}, + args: args{ + ctx: context.Background(), + data: []byte(fmt.Sprintf("You can find a xai secret %s within", secret)), + verify: true, + }, + want: []detectors.Result{ + { + DetectorType: detectorspb.DetectorType_XAI, + Verified: true, + }, + }, + wantErr: false, + wantVerificationErr: false, + }, + { + name: "found, unverified", + s: Scanner{}, + args: args{ + ctx: context.Background(), + data: []byte(fmt.Sprintf("You can find a xai secret %s within but not valid", inactiveSecret)), // the secret would satisfy the regex but not pass validation + verify: true, + }, + want: []detectors.Result{ + { + DetectorType: detectorspb.DetectorType_XAI, + Verified: false, + }, + }, + wantErr: false, + wantVerificationErr: false, + }, + { + name: "not found", + s: Scanner{}, + args: args{ + ctx: context.Background(), + data: []byte("You cannot find the secret within"), + verify: true, + }, + want: nil, + wantErr: false, + wantVerificationErr: false, + }, + { + name: "found, would be verified if not for timeout", + s: Scanner{client: common.SaneHttpClientTimeOut(1 * time.Microsecond)}, + args: args{ + ctx: context.Background(), + data: []byte(fmt.Sprintf("You can find a xai secret %s within", secret)), + verify: true, + }, + want: []detectors.Result{ + { + DetectorType: detectorspb.DetectorType_XAI, + Verified: false, + }, + }, + wantErr: false, + wantVerificationErr: true, + }, + { + name: "found, verified but unexpected api surface", + s: Scanner{client: common.ConstantResponseHttpClient(404, "")}, + args: args{ + ctx: context.Background(), + data: []byte(fmt.Sprintf("You can find a xai secret %s within", secret)), + verify: true, + }, + want: []detectors.Result{ + { + DetectorType: detectorspb.DetectorType_XAI, + Verified: false, + }, + }, + wantErr: false, + wantVerificationErr: true, + }, + } + for _, tt := range tests { + t.Run(tt.name, func(t *testing.T) { + got, err := tt.s.FromData(tt.args.ctx, tt.args.verify, tt.args.data) + if (err != nil) != tt.wantErr { + t.Errorf("Xai.FromData() error = %v, wantErr %v", err, tt.wantErr) + return + } + for i := range got { + if len(got[i].Raw) == 0 { + t.Fatalf("no raw secret present: \n %+v", got[i]) + } + if (got[i].VerificationError() != nil) != tt.wantVerificationErr { + t.Fatalf("wantVerificationError = %v, verification error = %v", tt.wantVerificationErr, got[i].VerificationError()) + } + } + ignoreOpts := cmpopts.IgnoreFields(detectors.Result{}, "Raw", "verificationError", "ExtraData") + if diff := cmp.Diff(got, tt.want, ignoreOpts); diff != "" { + t.Errorf("Xai.FromData() %s diff: (-got +want)\n%s", tt.name, diff) + } + }) + } +} + +func BenchmarkFromData(benchmark *testing.B) { + ctx := context.Background() + s := Scanner{} + for name, data := range detectors.MustGetBenchmarkData() { + benchmark.Run(name, func(b *testing.B) { + b.ResetTimer() + for n := 0; n < b.N; n++ { + _, err := s.FromData(ctx, false, data) + if err != nil { + b.Fatal(err) + } + } + }) + } +} diff --git a/pkg/detectors/xai/xai_test.go b/pkg/detectors/xai/xai_test.go new file mode 100644 index 000000000..3b8251c66 --- /dev/null +++ b/pkg/detectors/xai/xai_test.go @@ -0,0 +1,78 @@ +package xai + +import ( + "context" + "github.com/google/go-cmp/cmp" + "github.com/trufflesecurity/trufflehog/v3/pkg/detectors" + "github.com/trufflesecurity/trufflehog/v3/pkg/engine/ahocorasick" + "testing" +) + +func TestXai_Pattern(t *testing.T) { + d := Scanner{} + ahoCorasickCore := ahocorasick.NewAhoCorasickCore([]detectors.Detector{d}) + tests := []struct { + name string + input string + want []string + }{ + { + name: "typical pattern", + input: "xai_token = 'xai-W5zbfUkzlXedo7qD42AbBLlRSsyJrOW5zbfUkzlXedo7qD42AbBLlRSsyJrOW5zbfUkzlXedo7qD42Ab'", + want: []string{"xai-W5zbfUkzlXedo7qD42AbBLlRSsyJrOW5zbfUkzlXedo7qD42AbBLlRSsyJrOW5zbfUkzlXedo7qD42Ab"}, + }, + { + name: "finds all matches", + input: `grok_token1 = 'xai-W5zbfUkzlXedo7qD42AbBLlRSsyJrOW5zbfUkzlXedo7qD42AbBLlRSsyJrOW5zbfUkzlXedo7qD42Ab' +xai_token2 = 'xai-W5zbfUkzlXedo7qD42AbBLlRSsyJr1W5zbfUkzlXedo7qD42AbBLlRSsyJrOW5zbfUkzlXedo7qD42Ab'`, + want: []string{"xai-W5zbfUkzlXedo7qD42AbBLlRSsyJrOW5zbfUkzlXedo7qD42AbBLlRSsyJrOW5zbfUkzlXedo7qD42Ab", "xai-W5zbfUkzlXedo7qD42AbBLlRSsyJr1W5zbfUkzlXedo7qD42AbBLlRSsyJrOW5zbfUkzlXedo7qD42Ab"}, + }, + { + name: "invalid pattern", + input: "xai_token = 'xai-W5zbfUkzlXedo7qD42AbBLlRSsyJrOW5zbfUkzlXe'", + want: []string{}, + }, + } + + for _, test := range tests { + t.Run(test.name, func(t *testing.T) { + matchedDetectors := ahoCorasickCore.FindDetectorMatches([]byte(test.input)) + if len(matchedDetectors) == 0 { + t.Errorf("keywords '%v' not matched by: %s", d.Keywords(), test.input) + return + } + + results, err := d.FromData(context.Background(), false, []byte(test.input)) + if err != nil { + t.Errorf("error = %v", err) + return + } + + if len(results) != len(test.want) { + if len(results) == 0 { + t.Errorf("did not receive result") + } else { + t.Errorf("expected %d results, only received %d", len(test.want), len(results)) + } + return + } + + actual := make(map[string]struct{}, len(results)) + for _, r := range results { + if len(r.RawV2) > 0 { + actual[string(r.RawV2)] = struct{}{} + } else { + actual[string(r.Raw)] = struct{}{} + } + } + expected := make(map[string]struct{}, len(test.want)) + for _, v := range test.want { + expected[v] = struct{}{} + } + + if diff := cmp.Diff(expected, actual); diff != "" { + t.Errorf("%s diff: (-want +got)\n%s", test.name, diff) + } + }) + } +} diff --git a/pkg/engine/defaults/defaults.go b/pkg/engine/defaults/defaults.go index f1f8131a3..741816903 100644 --- a/pkg/engine/defaults/defaults.go +++ b/pkg/engine/defaults/defaults.go @@ -812,6 +812,7 @@ import ( "github.com/trufflesecurity/trufflehog/v3/pkg/detectors/worldcoinindex" "github.com/trufflesecurity/trufflehog/v3/pkg/detectors/worldweather" "github.com/trufflesecurity/trufflehog/v3/pkg/detectors/wrike" + "github.com/trufflesecurity/trufflehog/v3/pkg/detectors/xai" "github.com/trufflesecurity/trufflehog/v3/pkg/detectors/yandex" "github.com/trufflesecurity/trufflehog/v3/pkg/detectors/yelp" "github.com/trufflesecurity/trufflehog/v3/pkg/detectors/youneedabudget" @@ -1674,6 +1675,7 @@ func buildDetectorList() []detectors.Detector { &workstack.Scanner{}, &worldcoinindex.Scanner{}, &worldweather.Scanner{}, + &xai.Scanner{}, &wrike.Scanner{}, &yandex.Scanner{}, &yelp.Scanner{}, diff --git a/pkg/pb/detectorspb/detectors.pb.go b/pkg/pb/detectorspb/detectors.pb.go index 67cd1c348..4e4610330 100644 --- a/pkg/pb/detectorspb/detectors.pb.go +++ b/pkg/pb/detectorspb/detectors.pb.go @@ -1127,6 +1127,7 @@ const ( DetectorType_Harness DetectorType = 1020 DetectorType_Langfuse DetectorType = 1021 DetectorType_BingSubscriptionKey DetectorType = 1022 + DetectorType_XAI DetectorType = 1023 ) // Enum value maps for DetectorType. @@ -2151,6 +2152,7 @@ var ( 1020: "Harness", 1021: "Langfuse", 1022: "BingSubscriptionKey", + 1023: "XAI", } DetectorType_value = map[string]int32{ "Alibaba": 0, @@ -3172,6 +3174,7 @@ var ( "Harness": 1020, "Langfuse": 1021, "BingSubscriptionKey": 1022, + "XAI": 1023, } ) @@ -3625,7 +3628,7 @@ var file_detectors_proto_rawDesc = []byte{ 0x4c, 0x41, 0x49, 0x4e, 0x10, 0x01, 0x12, 0x0a, 0x0a, 0x06, 0x42, 0x41, 0x53, 0x45, 0x36, 0x34, 0x10, 0x02, 0x12, 0x09, 0x0a, 0x05, 0x55, 0x54, 0x46, 0x31, 0x36, 0x10, 0x03, 0x12, 0x13, 0x0a, 0x0f, 0x45, 0x53, 0x43, 0x41, 0x50, 0x45, 0x44, 0x5f, 0x55, 0x4e, 0x49, 0x43, 0x4f, 0x44, 0x45, - 0x10, 0x04, 0x2a, 0xd2, 0x83, 0x01, 0x0a, 0x0c, 0x44, 0x65, 0x74, 0x65, 0x63, 0x74, 0x6f, 0x72, + 0x10, 0x04, 0x2a, 0xdc, 0x83, 0x01, 0x0a, 0x0c, 0x44, 0x65, 0x74, 0x65, 0x63, 0x74, 0x6f, 0x72, 0x54, 0x79, 0x70, 0x65, 0x12, 0x0b, 0x0a, 0x07, 0x41, 0x6c, 0x69, 0x62, 0x61, 0x62, 0x61, 0x10, 0x00, 0x12, 0x08, 0x0a, 0x04, 0x41, 0x4d, 0x51, 0x50, 0x10, 0x01, 0x12, 0x07, 0x0a, 0x03, 0x41, 0x57, 0x53, 0x10, 0x02, 0x12, 0x09, 0x0a, 0x05, 0x41, 0x7a, 0x75, 0x72, 0x65, 0x10, 0x03, 0x12, @@ -4678,11 +4681,12 @@ var file_detectors_proto_rawDesc = []byte{ 0x07, 0x12, 0x0c, 0x0a, 0x07, 0x48, 0x61, 0x72, 0x6e, 0x65, 0x73, 0x73, 0x10, 0xfc, 0x07, 0x12, 0x0d, 0x0a, 0x08, 0x4c, 0x61, 0x6e, 0x67, 0x66, 0x75, 0x73, 0x65, 0x10, 0xfd, 0x07, 0x12, 0x18, 0x0a, 0x13, 0x42, 0x69, 0x6e, 0x67, 0x53, 0x75, 0x62, 0x73, 0x63, 0x72, 0x69, 0x70, 0x74, 0x69, - 0x6f, 0x6e, 0x4b, 0x65, 0x79, 0x10, 0xfe, 0x07, 0x42, 0x3d, 0x5a, 0x3b, 0x67, 0x69, 0x74, 0x68, - 0x75, 0x62, 0x2e, 0x63, 0x6f, 0x6d, 0x2f, 0x74, 0x72, 0x75, 0x66, 0x66, 0x6c, 0x65, 0x73, 0x65, - 0x63, 0x75, 0x72, 0x69, 0x74, 0x79, 0x2f, 0x74, 0x72, 0x75, 0x66, 0x66, 0x6c, 0x65, 0x68, 0x6f, - 0x67, 0x2f, 0x76, 0x33, 0x2f, 0x70, 0x6b, 0x67, 0x2f, 0x70, 0x62, 0x2f, 0x64, 0x65, 0x74, 0x65, - 0x63, 0x74, 0x6f, 0x72, 0x73, 0x70, 0x62, 0x62, 0x06, 0x70, 0x72, 0x6f, 0x74, 0x6f, 0x33, + 0x6f, 0x6e, 0x4b, 0x65, 0x79, 0x10, 0xfe, 0x07, 0x12, 0x08, 0x0a, 0x03, 0x58, 0x41, 0x49, 0x10, + 0xff, 0x07, 0x42, 0x3d, 0x5a, 0x3b, 0x67, 0x69, 0x74, 0x68, 0x75, 0x62, 0x2e, 0x63, 0x6f, 0x6d, + 0x2f, 0x74, 0x72, 0x75, 0x66, 0x66, 0x6c, 0x65, 0x73, 0x65, 0x63, 0x75, 0x72, 0x69, 0x74, 0x79, + 0x2f, 0x74, 0x72, 0x75, 0x66, 0x66, 0x6c, 0x65, 0x68, 0x6f, 0x67, 0x2f, 0x76, 0x33, 0x2f, 0x70, + 0x6b, 0x67, 0x2f, 0x70, 0x62, 0x2f, 0x64, 0x65, 0x74, 0x65, 0x63, 0x74, 0x6f, 0x72, 0x73, 0x70, + 0x62, 0x62, 0x06, 0x70, 0x72, 0x6f, 0x74, 0x6f, 0x33, } var ( diff --git a/proto/detectors.proto b/proto/detectors.proto index 06156ceb8..7247bdd7f 100644 --- a/proto/detectors.proto +++ b/proto/detectors.proto @@ -1032,6 +1032,7 @@ enum DetectorType { Harness = 1020; Langfuse = 1021; BingSubscriptionKey = 1022; + XAI = 1023; } message Result {