Identify transient AWS verification failures (#1563)

It turns out that GetCallerIdentity returns a surprising quantity of transient, false-negative 403 responses that carry the SignatureDoesNotMatch error reason. I don't know why this is happening, but their transient nature makes them indeterminate verification failures and they should be flagged as such. The AWS detector has therefore been modified to specifically look for the InvalidClientTokenId error reason in 403 responses and mark all other responses as indeterminate.

In addition to the functional changes this PR contains some updates to the test code that allow us to test them.

pkg/common/http.go

@@ -100,14 +100,14 @@ func NewCustomTransport(T http.RoundTripper) *CustomTransport {
 	return &CustomTransport{T}
 }
 
-func ConstantStatusHttpClient(statusCode int) *http.Client {
+func ConstantResponseHttpClient(statusCode int, body string) *http.Client {
 	return &http.Client{
 		Timeout: DefaultResponseTimeout,
 		Transport: FakeTransport{
 			CreateResponse: func(req *http.Request) (*http.Response, error) {
 				return &http.Response{
 					Request:    req,
-					Body:       io.NopCloser(strings.NewReader("")),
+					Body:       io.NopCloser(strings.NewReader(body)),
 					StatusCode: statusCode,
 				}, nil
 			},

pkg/detectors/alchemy/alchemy_test.go

@@ -104,7 +104,7 @@ func TestAlchemy_FromChunk(t *testing.T) {
 		},
 		{
 			name: "found, verified but unexpected api surface",
-			s:    Scanner{client: common.ConstantStatusHttpClient(404)},
+			s:    Scanner{client: common.ConstantResponseHttpClient(404, "")},
 			args: args{
 				ctx:    context.Background(),
 				data:   []byte(fmt.Sprintf("You can find a alchemy secret %s within", secret)),

pkg/detectors/aws/aws.go

@@ -18,7 +18,8 @@ import (
 )
 
 type scanner struct {
-	skipIDs map[string]struct{}
+	verificationClient *http.Client
+	skipIDs            map[string]struct{}
 }
 
 func New(opts ...func(*scanner)) *scanner {
@@ -48,7 +49,7 @@ func WithSkipIDs(skipIDs []string) func(*scanner) {
 var _ detectors.Detector = (*scanner)(nil)
 
 var (
-	client = common.SaneHttpClient()
+	defaultVerificationClient = common.SaneHttpClient()
 
 	// Make sure that your group is surrounded in boundary characters such as below to reduce false positives.
 	// Key types are from this list https://docs.aws.amazon.com/IAM/latest/UserGuide/reference_identifiers.html#identifiers-unique-ids
@@ -123,8 +124,9 @@ func (s scanner) FromData(ctx context.Context, verify bool, data []byte) (result
 				host := "sts.amazonaws.com"
 				region := "us-east-1"
 				endpoint := "https://sts.amazonaws.com"
-				datestamp := time.Now().UTC().Format("20060102")
-				amzDate := time.Now().UTC().Format("20060102T150405Z0700")
+				now := time.Now().UTC()
+				datestamp := now.Format("20060102")
+				amzDate := now.Format("20060102T150405Z0700")
 
 				req, err := http.NewRequestWithContext(ctx, method, endpoint, nil)
 				if err != nil {
@@ -171,6 +173,10 @@ func (s scanner) FromData(ctx context.Context, verify bool, data []byte) (result
 				req.Header.Add("Content-type", "application/x-www-form-urlencoded; charset=utf-8")
 				req.URL.RawQuery = params.Encode()
 
+				client := s.verificationClient
+				if client == nil {
+					client = defaultVerificationClient
+				}
 				res, err := client.Do(req)
 				if err == nil {
 
@@ -194,7 +200,25 @@ func (s scanner) FromData(ctx context.Context, verify bool, data []byte) (result
 							continue
 						}
 
-						if res.StatusCode != 403 {
+						if res.StatusCode == 403 {
+							var body awsErrorResponseBody
+							err = json.NewDecoder(res.Body).Decode(&body)
+							res.Body.Close()
+							if err == nil {
+								// All instances of the code I've seen in the wild are PascalCased but this check is
+								// case-insensitive out of an abundance of caution
+								if strings.EqualFold(body.Error.Code, "InvalidClientTokenId") {
+									// determinate failure - nothing to do
+								} else {
+									// We see a surprising number of false-negative signature mismatch errors here
+									// (The official SDK somehow elicits even more than just making the request
+									// ourselves)
+									s1.VerificationError = fmt.Errorf("request to %v returned status %d with an unexpected reason (%s: %s)", res.Request.URL, res.StatusCode, body.Error.Code, body.Error.Message)
+								}
+							} else {
+								s1.VerificationError = fmt.Errorf("couldn't parse the sts response body (%v)", err)
+							}
+						} else {
 							s1.VerificationError = fmt.Errorf("request to %v returned unexpected status %d", res.Request.URL, res.StatusCode)
 						}
 					}
@@ -245,6 +269,15 @@ func awsCustomCleanResults(results []detectors.Result) []detectors.Result {
 	return out
 }
 
+type awsError struct {
+	Code    string `json:"Code"`
+	Message string `json:"Message"`
+}
+
+type awsErrorResponseBody struct {
+	Error awsError `json:"Error"`
+}
+
 type identityRes struct {
 	GetCallerIdentityResponse struct {
 		GetCallerIdentityResult struct {

pkg/detectors/aws/aws_test.go

@@ -17,6 +17,8 @@ import (
 	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/detectorspb"
 )
 
+var unverifiedSecretClient = common.ConstantResponseHttpClient(403, `{"Error": {"Code": "InvalidClientTokenId"} }`)
+
 func TestAWS_FromChunk(t *testing.T) {
 	ctx, cancel := context.WithTimeout(context.Background(), time.Second*5)
 	defer cancel()
@@ -69,7 +71,7 @@ func TestAWS_FromChunk(t *testing.T) {
 		},
 		{
 			name: "found, unverified",
-			s:    scanner{},
+			s:    scanner{verificationClient: unverifiedSecretClient},
 			args: args{
 				ctx:    context.Background(),
 				data:   []byte(fmt.Sprintf("You can find a aws secret %s within aws %s but not valid", inactiveSecret, id)), // the secret would satisfy the regex but not pass validation
@@ -163,7 +165,7 @@ func TestAWS_FromChunk(t *testing.T) {
 		},
 		{
 			name: "found, unverified, with leading +",
-			s:    scanner{},
+			s:    scanner{verificationClient: unverifiedSecretClient},
 			args: args{
 				ctx:    context.Background(),
 				data:   []byte(fmt.Sprintf("You can find a aws secret %s within aws %s but not valid", "+HaNv9cTwheDKGJaws/+BMF2GgybQgBWdhcOOdfF", id)), // the secret would satisfy the regex but not pass validation
@@ -194,9 +196,45 @@ func TestAWS_FromChunk(t *testing.T) {
 		},
 		{
 			name: "found, would be verified if not for http timeout",
-			s:    scanner{},
+			s:    scanner{verificationClient: common.SaneHttpClientTimeOut(1 * time.Microsecond)},
 			args: args{
-				ctx:    timeoutContext(1 * time.Microsecond),
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a aws secret %s within aws %s", secret, id)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_AWS,
+					Verified:     false,
+					Redacted:     "AKIASP2TPHJSQH3FJRUX",
+				},
+			},
+			wantErr:               false,
+			wantVerificationError: true,
+		},
+		{
+			name: "found, unverified due to unexpected http response status",
+			s:    scanner{verificationClient: common.ConstantResponseHttpClient(500, "internal server error")},
+			args: args{
+				ctx:    context.Background(),
+				data:   []byte(fmt.Sprintf("You can find a aws secret %s within aws %s", secret, id)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_AWS,
+					Verified:     false,
+					Redacted:     "AKIASP2TPHJSQH3FJRUX",
+				},
+			},
+			wantErr:               false,
+			wantVerificationError: true,
+		},
+		{
+			name: "found, unverified due to unexpected 403 response reason",
+			s:    scanner{verificationClient: common.ConstantResponseHttpClient(403, `{"Error": {"Code": "SignatureDoesNotMatch"} }`)},
+			args: args{
+				ctx:    context.Background(),
 				data:   []byte(fmt.Sprintf("You can find a aws secret %s within aws %s", secret, id)),
 				verify: true,
 			},
@@ -224,7 +262,7 @@ func TestAWS_FromChunk(t *testing.T) {
 					t.Fatalf("no raw secret present: \n %+v", got[i])
 				}
 				if (got[i].VerificationError != nil) != tt.wantVerificationError {
-					t.Fatalf("verification error = %v, wantVerificationError %v", got[i].VerificationError, tt.wantVerificationError)
+					t.Fatalf("wantVerificationError %v, verification error = %v", tt.wantVerificationError, got[i].VerificationError)
 				}
 			}
 			ignoreOpts := cmpopts.IgnoreFields(detectors.Result{}, "RawV2", "Raw", "VerificationError")
@@ -249,8 +287,3 @@ func BenchmarkFromData(benchmark *testing.B) {
 		})
 	}
 }
-
-func timeoutContext(timeout time.Duration) context.Context {
-	c, _ := context.WithTimeout(context.Background(), timeout)
-	return c
-}