[INS-406] Braintrust detector

pkg/detectors/braintrust/braintrust.go

@@ -0,0 +1,133 @@
+package braintrust
+
+import (
+	"context"
+	"fmt"
+	"io"
+	"net/http"
+
+	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
+	detectors.DefaultMultiPartCredentialProvider
+}
+
+// Compile-time interface check
+var _ detectors.Detector = (*Scanner)(nil)
+
+var (
+	defaultClient = common.SaneHttpClient()
+
+	// Braintrust API keys:
+	// Format: sk- + 48 alphanumeric characters (observed)
+	braintrustTokenPat = regexp.MustCompile(
+		`\b(sk-[A-Za-z0-9]{48})\b`,
+	)
+)
+
+// Keywords used for fast pre-filtering
+func (s Scanner) Keywords() []string {
+	return []string{"braintrust"}
+}
+
+func (s Scanner) getClient() *http.Client {
+	if s.client != nil {
+		return s.client
+	}
+	return defaultClient
+}
+
+// FromData scans for Braintrust API tokens and optionally verifies them
+func (s Scanner) FromData(
+	ctx context.Context,
+	verify bool,
+	data []byte,
+) (results []detectors.Result, err error) {
+
+	dataStr := string(data)
+
+	uniqueTokens := make(map[string]struct{})
+	for _, match := range braintrustTokenPat.FindAllStringSubmatch(dataStr, -1) {
+		uniqueTokens[match[1]] = struct{}{}
+	}
+
+	for token := range uniqueTokens {
+		result := detectors.Result{
+			DetectorType: detectorspb.DetectorType_BrainTrustApiKey,
+			Raw:          []byte(token),
+			Redacted:     token[:8] + "...",
+		}
+
+		if verify {
+			verified, verificationErr := verifyBraintrustToken(
+				ctx,
+				s.getClient(),
+				token,
+			)
+			result.SetVerificationError(verificationErr, token)
+			result.Verified = verified
+		}
+
+		results = append(results, result)
+	}
+
+	return
+}
+
+func verifyBraintrustToken(
+	ctx context.Context,
+	client *http.Client,
+	token string,
+) (bool, error) {
+
+	req, err := http.NewRequestWithContext(
+		ctx,
+		http.MethodGet,
+		"https://api.braintrust.dev/v1/project?limit=1",
+		http.NoBody,
+	)
+	if err != nil {
+		return false, err
+	}
+
+	req.Header.Set("Authorization", "Bearer "+token)
+
+	res, err := client.Do(req)
+	if err != nil {
+		return false, err
+	}
+	defer func() {
+		_, _ = io.Copy(io.Discard, res.Body)
+		_ = res.Body.Close()
+	}()
+
+	switch res.StatusCode {
+	case http.StatusOK:
+		return true, nil
+	case http.StatusUnauthorized:
+		// Invalid or revoked token
+		return false, nil
+	case http.StatusForbidden:
+		// Valid token but lacks permission (still valid)
+		return true, nil
+	default:
+		return false, fmt.Errorf(
+			"unexpected HTTP response status %d",
+			res.StatusCode,
+		)
+	}
+}
+
+func (s Scanner) Type() detectorspb.DetectorType {
+	return detectorspb.DetectorType_BrainTrustApiKey
+}
+
+func (s Scanner) Description() string {
+	return "Braintrust is an AI evaluation and observability platform. This detector identifies Braintrust API keys."
+}

pkg/detectors/braintrust/braintrust_integration_test.go

@@ -0,0 +1,178 @@
+//go:build detectors
+// +build detectors
+
+package braintrust
+
+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 TestBraintrust_FromData(t *testing.T) {
+	ctx, cancel := context.WithTimeout(context.Background(), time.Second*10)
+	defer cancel()
+
+	// Load secrets from GCP (same pattern as other detectors)
+	testSecrets, err := common.GetSecret(ctx, "trufflehog-testing", "detectors6")
+	if err != nil {
+		t.Fatalf("could not get test secrets from GCP: %s", err)
+	}
+
+	activeToken := testSecrets.MustGetField("BRAINTRUST_API_KEY")
+	inactiveToken := "sk-AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA"
+
+	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:   fmt.Appendf([]byte{}, "Using Braintrust API key %s", activeToken),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_BrainTrustApiKey,
+					Verified:     true,
+					Raw:          []byte(activeToken),
+					Redacted:     activeToken[:8] + "...",
+				},
+			},
+		},
+		{
+			name: "found, real token, verification error due to timeout",
+			s:    Scanner{client: common.SaneHttpClientTimeOut(1 * time.Microsecond)},
+			args: args{
+				ctx:    context.Background(),
+				data:   fmt.Appendf([]byte{}, "Using Braintrust API key %s", activeToken),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_BrainTrustApiKey,
+					Verified:     false,
+					Raw:          []byte(activeToken),
+					Redacted:     activeToken[:8] + "...",
+				},
+			},
+			wantVerificationErr: true,
+		},
+		{
+			name: "found, real token, verification error due to unexpected api surface",
+			s:    Scanner{client: common.ConstantResponseHttpClient(500, "{}")},
+			args: args{
+				ctx:    context.Background(),
+				data:   fmt.Appendf([]byte{}, "Using Braintrust API key %s", activeToken),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_BrainTrustApiKey,
+					Verified:     false,
+					Raw:          []byte(activeToken),
+					Redacted:     activeToken[:8] + "...",
+				},
+			},
+			wantVerificationErr: true,
+		},
+		{
+			name: "found, unverified (inactive token)",
+			s:    Scanner{},
+			args: args{
+				ctx:    context.Background(),
+				data:   fmt.Appendf([]byte{}, "Using Braintrust API key %s", inactiveToken),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_BrainTrustApiKey,
+					Verified:     false,
+					Raw:          []byte(inactiveToken),
+					Redacted:     inactiveToken[:8] + "...",
+				},
+			},
+		},
+		{
+			name: "not found",
+			s:    Scanner{},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte("no secrets here"),
+				verify: true,
+			},
+			want: nil,
+		},
+	}
+
+	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.Fatalf("Braintrust.FromData() error = %v, wantErr %v", err, tt.wantErr)
+			}
+
+			for i := range got {
+				if len(got[i].Raw) == 0 {
+					t.Fatal("no raw secret present")
+				}
+				if (got[i].VerificationError() != nil) != tt.wantVerificationErr {
+					t.Fatalf(
+						"wantVerificationError = %v, verification error = %v",
+						tt.wantVerificationErr,
+						got[i].VerificationError(),
+					)
+				}
+			}
+
+			ignoreOpts := cmpopts.IgnoreFields(
+				detectors.Result{},
+				"ExtraData",
+				"verificationError",
+				"primarySecret",
+			)
+
+			if diff := cmp.Diff(got, tt.want, ignoreOpts); diff != "" {
+				t.Errorf("Braintrust.FromData() %s diff: (-got +want)\n%s", tt.name, diff)
+			}
+		})
+	}
+}
+
+func BenchmarkBraintrust_FromData(b *testing.B) {
+	ctx := context.Background()
+	s := Scanner{}
+
+	for name, data := range detectors.MustGetBenchmarkData() {
+		b.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)
+				}
+			}
+		})
+	}
+}