trufflesecurity/trufflehog
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Detector-Competition-New: add anthropic api key detector (#1861)

Browse Source 
* feat(anthropic): add anthropic api key detector

* Detector-Competition-Fix: fix remove debug println
This commit is contained in:
Corben Leo
2023-10-05 11:34:40 -05:00
committed by GitHub
parent bf1cce43e5
commit 179a7e4cbc
5 changed files with 288 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
pkg/detectors/anthropic/anthropic.go
+118
View File
@@ -0,0 +1,118 @@
package anthropic
import (
"bytes"
"context"
"encoding/json"
"fmt"
"net/http"
"regexp"
"strings"
"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(sk-ant-api03-[\w\-]{93}AA)\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{"sk-ant-api03"}
}
type response struct {
Error struct {
Type string `json:"type"`
Message string `json:"message"`
} `json:"error"`
}
// FromData will find and optionally verify Anthropic 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)
matches := keyPat.FindAllStringSubmatch(dataStr, -1)
for _, match := range matches {
if len(match) != 2 {
continue
}
resMatch := strings.TrimSpace(match[1])
s1 := detectors.Result{
DetectorType: detectorspb.DetectorType_Anthropic,
Raw: []byte(resMatch),
}
if verify {
client := s.client
if client == nil {
client = defaultClient
}
body := map[string]interface{}{
"prompt": "test",
}
bodyBytes, err := json.Marshal(body)
if err != nil {
continue
}
req, err := http.NewRequestWithContext(ctx, "POST", "https://api.anthropic.com/v1/complete", bytes.NewReader(bodyBytes))
if err != nil {
continue
}
req.Header.Set("x-api-key", resMatch)
req.Header.Set("Content-Type", "application/json")
res, err := client.Do(req)
if err == nil {
defer res.Body.Close()
if res.StatusCode == http.StatusBadRequest {
var resp response
if err = json.NewDecoder(res.Body).Decode(&resp); err != nil {
s1.VerificationError = fmt.Errorf("unexpected HTTP response body: %w", err)
continue
}
if resp.Error.Message == "max_tokens_to_sample: field required" {
// The secret is verified
// Anthropic returns 400 on a request containing a valid API key,
// when not containing a valid model field
s1.Verified = true
}
} else if res.StatusCode == 401 {
// The secret is determinately not verified (nothing to do)
// Anthropic returns 401 on all requests not containing a valid x-api-key header
} else {
s1.VerificationError = fmt.Errorf("unexpected HTTP response status %d", res.StatusCode)
}
} else {
s1.VerificationError = err
}
}
// This function will check false positives for common test words, but also it will make sure the key appears 'random' enough to be a real key.
if !s1.Verified && detectors.IsKnownFalsePositive(resMatch, detectors.DefaultFalsePositives, true) {
continue
}
results = append(results, s1)
}
return results, nil
}
func (s Scanner) Type() detectorspb.DetectorType {
return detectorspb.DetectorType_Anthropic
}
pkg/detectors/anthropic/anthropic_test.go
+161
View File
@@ -0,0 +1,161 @@
//go:build detectors
// +build detectors
package anthropic
import (
"context"
"fmt"
"github.com/google/go-cmp/cmp"
"github.com/google/go-cmp/cmp/cmpopts"
"testing"
"time"
"github.com/trufflesecurity/trufflehog/v3/pkg/detectors"
"github.com/trufflesecurity/trufflehog/v3/pkg/common"
"github.com/trufflesecurity/trufflehog/v3/pkg/pb/detectorspb"
)
func TestAnthropic_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("ANTHROPIC")
inactiveSecret := testSecrets.MustGetField("ANTHROPIC_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 anthropic secret %s within", secret)),
verify: true,
},
want: []detectors.Result{
{
DetectorType: detectorspb.DetectorType_Anthropic,
Verified: true,
},
},
wantErr: false,
wantVerificationErr: false,
},
{
name: "found, unverified",
s: Scanner{},
args: args{
ctx: context.Background(),
data: []byte(fmt.Sprintf("You can find a anthropic 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_Anthropic,
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 anthropic secret %s within", secret)),
verify: true,
},
want: []detectors.Result{
{
DetectorType: detectorspb.DetectorType_Anthropic,
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 anthropic secret %s within", secret)),
verify: true,
},
want: []detectors.Result{
{
DetectorType: detectorspb.DetectorType_Anthropic,
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("Anthropic.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("Anthropic.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)
}
}
})
}
}
pkg/engine/defaults.go
+2
View File
@@ -1,6 +1,7 @@
package engine
import (
"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/anthropic"
"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/envoyapikey"
"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/huggingface"
"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/ramp"
@@ -1550,6 +1551,7 @@ func DefaultDetectors() []detectors.Detector {
tailscale.Scanner{},
web3storage.Scanner{},
&ramp.Scanner{},
&anthropic.Scanner{},
}
}
pkg/pb/detectorspb/detectors.pb.go
+6 -2
View File
@@ -1004,6 +1004,7 @@ const (
DetectorType_Web3Storage DetectorType = 930
DetectorType_AzureStorage DetectorType = 931
DetectorType_PlanetScaleDb DetectorType = 932
DetectorType_Anthropic DetectorType = 933
DetectorType_Ramp DetectorType = 934
)
@@ -1939,6 +1940,7 @@ var (
930: "Web3Storage",
931: "AzureStorage",
932: "PlanetScaleDb",
933: "Anthropic",
934: "Ramp",
}
DetectorType_value = map[string]int32{
@@ -2871,6 +2873,7 @@ var (
"Web3Storage": 930,
"AzureStorage": 931,
"PlanetScaleDb": 932,
"Anthropic": 933,
"Ramp": 934,
}
)
@@ -3250,7 +3253,7 @@ var file_detectors_proto_rawDesc = []byte{
0x44, 0x65, 0x63, 0x6f, 0x64, 0x65, 0x72, 0x54, 0x79, 0x70, 0x65, 0x12, 0x0b, 0x0a, 0x07, 0x55,
0x4e, 0x4b, 0x4e, 0x4f, 0x57, 0x4e, 0x10, 0x00, 0x12, 0x09, 0x0a, 0x05, 0x50, 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, 0x2a, 0xf2, 0x74, 0x0a, 0x0c, 0x44,
0x09, 0x0a, 0x05, 0x55, 0x54, 0x46, 0x31, 0x36, 0x10, 0x03, 0x2a, 0x82, 0x75, 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,
@@ -4185,7 +4188,8 @@ var file_detectors_proto_rawDesc = []byte{
0x65, 0x62, 0x33, 0x53, 0x74, 0x6f, 0x72, 0x61, 0x67, 0x65, 0x10, 0xa2, 0x07, 0x12, 0x11, 0x0a,
0x0c, 0x41, 0x7a, 0x75, 0x72, 0x65, 0x53, 0x74, 0x6f, 0x72, 0x61, 0x67, 0x65, 0x10, 0xa3, 0x07,
0x12, 0x12, 0x0a, 0x0d, 0x50, 0x6c, 0x61, 0x6e, 0x65, 0x74, 0x53, 0x63, 0x61, 0x6c, 0x65, 0x44,
0x62, 0x10, 0xa4, 0x07, 0x12, 0x09, 0x0a, 0x04, 0x52, 0x61, 0x6d, 0x70, 0x10, 0xa6, 0x07, 0x42,
0x62, 0x10, 0xa4, 0x07, 0x12, 0x0e, 0x0a, 0x09, 0x41, 0x6e, 0x74, 0x68, 0x72, 0x6f, 0x70, 0x69,
0x63, 0x10, 0xa5, 0x07, 0x12, 0x09, 0x0a, 0x04, 0x52, 0x61, 0x6d, 0x70, 0x10, 0xa6, 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,
proto/detectors.proto
+1
View File
@@ -941,6 +941,7 @@ enum DetectorType {
Web3Storage = 930;
AzureStorage = 931;
PlanetScaleDb = 932;
Anthropic = 933;
Ramp = 934;
}