Elevenlabs detector (#3023)

* adding v1 eleven labs

* updating elevenlabs to support old and new version

* fixing status codes

* lint fixes

* adding test for v2

* adding test for v1

* return err

---------

Co-authored-by: Dylan Ayrey <dylan@Dylans-MacBook-Pro.local>
Co-authored-by: āh̳̕mͭͭͨͩ̐e̘ͬ́͋ͬ̊̓͂d <13666360+0x1@users.noreply.github.com>
This commit is contained in:
dylanTruffle 2024-07-03 07:53:20 -07:00 committed by GitHub
parent b883645856
commit 47535830c4
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
7 changed files with 569 additions and 6 deletions

View file

@ -0,0 +1,116 @@
package elevenlabs
import (
"context"
"encoding/json"
"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
}
func (Scanner) Version() int { return 1 }
type UserRes struct {
Subscription struct {
Tier string `json:"tier"`
} `json:"subscription"`
Name string `json:"first_name"`
}
// 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(`(?i)(?:elevenlabs|xi-api-key|el|token|key)[^\.].{0,40}[ =:'"]+([a-f0-9]{32})\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{"elevenlabs", "xi-api-key", "xi_api_key"}
}
// FromData will find and optionally verify Elevenlabs 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)
uniqueMatches := make(map[string]struct{})
for _, match := range keyPat.FindAllStringSubmatch(dataStr, -1) {
uniqueMatches[match[1]] = struct{}{}
}
for match := range uniqueMatches {
s1 := detectors.Result{
DetectorType: detectorspb.DetectorType_ElevenLabs,
Raw: []byte(match),
ExtraData: map[string]string{"version": "1"},
}
if verify {
client := s.client
if client == nil {
client = defaultClient
}
isVerified, userResponse, verificationErr := verifyMatch(ctx, client, match)
s1.Verified = isVerified
if userResponse != nil {
s1.ExtraData["Name"] = userResponse.Name
s1.ExtraData["Tier"] = userResponse.Subscription.Tier
}
s1.SetVerificationError(verificationErr, match)
}
results = append(results, s1)
}
return
}
func verifyMatch(ctx context.Context, client *http.Client, token string) (bool, *UserRes, error) {
req, err := http.NewRequestWithContext(ctx, http.MethodGet, "https://api.elevenlabs.io/v1/user", nil)
if err != nil {
return false, nil, err
}
req.Header.Set("xi-api-key", token)
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:
// If the endpoint returns useful information, we can return it as a map.
var userResponse UserRes
if err = json.NewDecoder(res.Body).Decode(&userResponse); err != nil {
return false, nil, err
}
return true, &userResponse, 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_ElevenLabs
}

View file

@ -0,0 +1,88 @@
//go:build detectors
// +build detectors
package elevenlabs
import (
"context"
"testing"
"github.com/google/go-cmp/cmp"
"github.com/trufflesecurity/trufflehog/v3/pkg/detectors"
"github.com/trufflesecurity/trufflehog/v3/pkg/engine/ahocorasick"
)
func TestElevenlabs_Pattern(t *testing.T) {
d := Scanner{}
ahoCorasickCore := ahocorasick.NewAhoCorasickCore([]detectors.Detector{d})
tests := []struct {
name string
input string
want []string
}{
{
name: "typical pattern",
input: "XI_API_KEY = 'b41b9d78aefb8c7c6cf9ebf01231340b'",
want: []string{"b41b9d78aefb8c7c6cf9ebf01231340b"},
},
}
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)
}
})
}
}
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)
}
}
})
}
}

View file

@ -0,0 +1,116 @@
package elevenlabs
import (
"context"
"encoding/json"
"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
}
func (Scanner) Version() int { return 2 }
type UserRes struct {
Subscription struct {
Tier string `json:"tier"`
} `json:"subscription"`
Name string `json:"first_name"`
}
// 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((?:sk)_[a-f0-9]{48})\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{"elevenlabs", "xi-api-key", "el", "xi_api_key"}
}
// FromData will find and optionally verify Elevenlabs 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)
uniqueMatches := make(map[string]struct{})
for _, match := range keyPat.FindAllStringSubmatch(dataStr, -1) {
uniqueMatches[match[1]] = struct{}{}
}
for match := range uniqueMatches {
s1 := detectors.Result{
DetectorType: detectorspb.DetectorType_ElevenLabs,
Raw: []byte(match),
ExtraData: map[string]string{"version": "2"},
}
if verify {
client := s.client
if client == nil {
client = defaultClient
}
isVerified, userResponse, verificationErr := verifyMatch(ctx, client, match)
s1.Verified = isVerified
if userResponse != nil {
s1.ExtraData["Name"] = userResponse.Name
s1.ExtraData["Tier"] = userResponse.Subscription.Tier
}
s1.SetVerificationError(verificationErr, match)
}
results = append(results, s1)
}
return
}
func verifyMatch(ctx context.Context, client *http.Client, token string) (bool, *UserRes, error) {
req, err := http.NewRequestWithContext(ctx, http.MethodGet, "https://api.elevenlabs.io/v1/user", nil)
if err != nil {
return false, nil, err
}
req.Header.Set("xi-api-key", token)
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:
// If the endpoint returns useful information, we can return it as a map.
var userResponse UserRes
if err = json.NewDecoder(res.Body).Decode(&userResponse); err != nil {
return false, nil, err
}
return true, &userResponse, 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_ElevenLabs
}

View file

@ -0,0 +1,234 @@
//go:build detectors
// +build detectors
package elevenlabs
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/engine/ahocorasick"
"github.com/trufflesecurity/trufflehog/v3/pkg/pb/detectorspb"
)
func TestElevenlabs_Pattern(t *testing.T) {
d := Scanner{}
ahoCorasickCore := ahocorasick.NewAhoCorasickCore([]detectors.Detector{d})
tests := []struct {
name string
input string
want []string
}{
{
name: "typical pattern",
input: "XI_API_KEY = 'sk_c43667f9bedd46fcff858f09f648d984533645e30f0541df'",
want: []string{"sk_c43667f9bedd46fcff858f09f648d984533645e30f0541df"},
},
}
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)
}
})
}
}
func TestElevenlabs_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("ELEVENLABS")
inactiveSecret := testSecrets.MustGetField("ELEVENLABS_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 elevenlabs secret %s within", secret)),
verify: true,
},
want: []detectors.Result{
{
DetectorType: detectorspb.DetectorType_ElevenLabs,
Verified: true,
ExtraData: map[string]string{
"version": "2",
"Name": "Ahmed",
"Tier": "free",
},
},
},
wantErr: false,
wantVerificationErr: false,
},
{
name: "found, unverified",
s: Scanner{},
args: args{
ctx: context.Background(),
data: []byte(fmt.Sprintf("You can find a elevenlabs 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_ElevenLabs,
Verified: false,
ExtraData: map[string]string{
"version": "2",
},
},
},
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 elevenlabs secret %s within", secret)),
verify: true,
},
want: []detectors.Result{
{
DetectorType: detectorspb.DetectorType_ElevenLabs,
Verified: false,
ExtraData: map[string]string{
"version": "2",
},
},
},
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 elevenlabs secret %s within", secret)),
verify: true,
},
want: []detectors.Result{
{
DetectorType: detectorspb.DetectorType_ElevenLabs,
Verified: false,
ExtraData: map[string]string{
"version": "2",
},
},
},
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("Elevenlabs.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")
if diff := cmp.Diff(got, tt.want, ignoreOpts); diff != "" {
t.Errorf("Elevenlabs.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)
}
}
})
}
}

View file

@ -230,6 +230,8 @@ import (
"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/edenai"
"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/eightxeight"
"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/elasticemail"
elevenlabsv1 "github.com/trufflesecurity/trufflehog/v3/pkg/detectors/elevenlabs/v1"
elevenlabsv2 "github.com/trufflesecurity/trufflehog/v3/pkg/detectors/elevenlabs/v2"
"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/enablex"
"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/endorlabs"
"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/enigma"
@ -1042,6 +1044,8 @@ func DefaultDetectors() []detectors.Detector {
&prospectcrm.Scanner{},
&surveyanyplace.Scanner{},
&ubidots.Scanner{},
&elevenlabsv1.Scanner{},
&elevenlabsv2.Scanner{},
sinchmessage.Scanner{},
ayrshare.Scanner{},
mailboxlayer.Scanner{},

View file

@ -1093,6 +1093,7 @@ const (
DetectorType_LarkSuite DetectorType = 991
DetectorType_LarkSuiteApiKey DetectorType = 992
DetectorType_EndorLabs DetectorType = 993
DetectorType_ElevenLabs DetectorType = 994
)
// Enum value maps for DetectorType.
@ -2088,6 +2089,7 @@ var (
991: "LarkSuite",
992: "LarkSuiteApiKey",
993: "EndorLabs",
994: "ElevenLabs",
}
DetectorType_value = map[string]int32{
"Alibaba": 0,
@ -3080,6 +3082,7 @@ var (
"LarkSuite": 991,
"LarkSuiteApiKey": 992,
"EndorLabs": 993,
"ElevenLabs": 994,
}
)
@ -3533,7 +3536,7 @@ 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, 0xf0, 0x7e, 0x0a, 0x0c, 0x44, 0x65, 0x74, 0x65, 0x63, 0x74, 0x6f, 0x72, 0x54,
0x10, 0x04, 0x2a, 0x81, 0x7f, 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, 0x0a,
@ -4548,11 +4551,12 @@ var file_detectors_proto_rawDesc = []byte{
0x0e, 0x0a, 0x09, 0x4c, 0x61, 0x72, 0x6b, 0x53, 0x75, 0x69, 0x74, 0x65, 0x10, 0xdf, 0x07, 0x12,
0x14, 0x0a, 0x0f, 0x4c, 0x61, 0x72, 0x6b, 0x53, 0x75, 0x69, 0x74, 0x65, 0x41, 0x70, 0x69, 0x4b,
0x65, 0x79, 0x10, 0xe0, 0x07, 0x12, 0x0e, 0x0a, 0x09, 0x45, 0x6e, 0x64, 0x6f, 0x72, 0x4c, 0x61,
0x62, 0x73, 0x10, 0xe1, 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,
0x62, 0x73, 0x10, 0xe1, 0x07, 0x12, 0x0f, 0x0a, 0x0a, 0x45, 0x6c, 0x65, 0x76, 0x65, 0x6e, 0x4c,
0x61, 0x62, 0x73, 0x10, 0xe2, 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 (

View file

@ -1003,6 +1003,7 @@ enum DetectorType {
LarkSuite = 991;
LarkSuiteApiKey = 992;
EndorLabs = 993;
ElevenLabs = 994;
}
message Result {