Back to skill

Security audit

didit-sdk-integration

Security checks for vulnerabilities and agentic risk

Overview

ClawScan marks this Didit SDK integration skill benign because the reported network calls and credential use appear aligned with SDK integration, though the scan was only partial and confidence is medium.

The verdict is benign, not risk-free. Consider the incomplete scan before installing: Skillspector reports 50% coverage, one static analyzer was degraded, and AIG reported no findings. Review how the skill uses your API key and which data it sends to Didit, check the unpinned Expo references, and confirm the host’s tool permissions.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Rogue AgentSelf-Modification, Session Persistence
  • Taint TrackingDirect Taint Flow, Variable-Mediated Taint Flow, Credential Exfiltration Chain
Findings (15)

Tainted flow: 'api_key' from os.environ.get (line 118, credential/environment) → requests.post (network output)

Critical
Category
Data Flow
Confidence
90% confidence
Finding

Credentials or environment variables flow to a network sink. This is a high-confidence indicator of credential exfiltration.

Content

Scanner excerpt · scripts/mint_sdk_tokens.py (reported line 57)May include surrounding context.

python
def post(path: str, payload: dict, api_key: str) -> dict:
    response = requests.post(
        f"{BASE_URL}{path}",
        headers={"x-api-key": api_key, "Content-Type": "application/json"},
        json=payload,

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · SKILL.md (reported line 156)May include surrounding context.

md
runtime URL and exact registered redirects/origins. The documented early-access
hosted sign-in/ flow uses random state and S256 PKCE, handles configured sign-up,
login, reset and MFA, and returns a code. Check callback state, then exchange the
code at token/ with the original verifier and redirect URI. The access token format
depends on the application's Authentication runtime as documented: read the current
tokens page before choosing. An opaque token is validated via GET session/ (never
decoded as a JWT); an ES256 JWT is verified against the issuer's published JWKS.

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · SKILL.md (reported line 171)May include surrounding context.

md
https://docs.didit.me/authentication/events-and-webhooks for the release gates.
For a JWT runtime, use a trusted configured issuer/JWKS, validate signature, algorithm,
issuer, audience and time claims, and use a live session check when immediate
revocation matters. Never invent a JWKS URL. An ID token is not an API access token.

Test the selected products alone and together using one server-resolved vendor_data.
Cover cross-user/application rejection, callback state/replay, expiry, refresh races,

Skill declares no tool scope ('permissions' or 'allowed-tools') but code capabilities were detected: env, network.

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 133)May include surrounding context.

md
the actual decision. Sign-in success alone never approves identity verification.

Transaction monitoring, only if selected:
No verification workflow is needed just to submit transactions. Mint
POST /v3/transactions/sdk-token/ on the backend for the user's vendor_data and
return sdk_token to submitTransaction. Use the platform's actual payload schema:
web category/details differ from mobile type/info and the REST wire schema.

External Transmission

Medium
Category
Data Exfiltration
Confidence
70% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 225)May include surrounding context.

md
import requests

def create_didit_session(user_id: str) -> dict:
    response = requests.post(
        "https://verification.didit.me/v3/session/",
        headers={"x-api-key": os.environ["DIDIT_API_KEY"], "Content-Type": "application/json"},
        json={

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 225)May include surrounding context.

md
import requests

def create_didit_session(user_id: str) -> dict:
    response = requests.post(
        "https://verification.didit.me/v3/session/",
        headers={"x-api-key": os.environ["DIDIT_API_KEY"], "Content-Type": "application/json"},
        json={

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 245)May include surrounding context.

TypeScript (Node backend):

typescript
const response = await fetch("https://verification.didit.me/v3/session/", {
  method: "POST",
  headers: { "x-api-key": process.env.DIDIT_API_KEY!, "Content-Type": "application/json" },
  body: JSON.stringify({ workflow_id: process.env.DIDIT_WORKFLOW_ID, vendor_data: userId }),

Session Persistence

Medium
Category
Rogue Agent
Confidence
75% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · SKILL.md (reported line 356)May include surrounding context.

SPM does not enforce the iOS 15 floor: set the deployment target to 15.0 yourself for DiditSDK or DiditSDKNFC. Code always writes import DiditSDK.

Info.plist: NSCameraUsageDescription, NSMicrophoneUsageDescription and NSPhotoLibraryUsageDescription are required (iOS terminates the app when a missing key's resource is accessed). With the NFC or All variant also add NFCReaderUsageDescription, the Near Field Communication Tag Reading capability, the entitlement com.apple.developer.nfc.readersession.formats = TAG, and com.apple.developer.nfc.readersession.iso7816.select-identifiers = D23300000045737445494420763335, A0000002471001, A0000002472001, 00000000000000. NFC does not run on the simulator.

swift
import SwiftUI

Session Persistence

Medium
Category
Rogue Agent
Confidence
75% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · SKILL.md (reported line 450)May include surrounding context.

text

- Expo config plugin options: `{ "iosNfcEnabled": false, "androidNfcEnabled": false }` builds without NFC. The plugin adds the Maven repository, the podspec and the Android packaging rules.
- React Native CLI: add the DiditSDK podspec pinned to tag `4.9.1` (never `main`) to the Podfile, the Didit Maven repository to `settings.gradle`, and the packaging rules from the docs; `DIDIT_SDK_IOS_NFC_ENABLED=false` / `diditSdkAndroidNfcEnabled=false` turn NFC off. Add the iOS Info.plist keys listed above.

```typescript
import { startVerification, VerificationStatus } from "@didit-protocol/sdk-react-native";

Session Persistence

Medium
Category
Rogue Agent
Confidence
75% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · SKILL.md (reported line 478)May include surrounding context.

flutter pub add didit_sdk

text

iOS: select the native variant in `ios/Podfile` (`$DiditSdkIosVariant = 'all'`, `'core'`, `'autodetection'` or `'nfc'`; `all` and `nfc` need `platform :ios, '15.0'`) with the podspec pinned to tag `4.9.1`, then add the Info.plist keys listed above. Android: select the native variant with `diditSdkAndroidVariant` in `android/gradle.properties` (`all` by default, `core`, `autodetection` or `nfc`; `core` and `autodetection` drop NFC), or depend on `didit_sdk_core` / `didit_sdk_autodetection` / `didit_sdk_nfc` instead of `didit_sdk` to pin the variant on both platforms (then import from that package, for example `package:didit_sdk_core/sdk_flutter.dart`). `all` and `nfc` builds need `pickFirsts += "META-INF/versions/9/OSGI-INF/MANIFEST.MF"` in `packaging.resources`. Full snippets: https://docs.didit.me/integration/native-sdks/flutter-sdk

```dart
import 'package:didit_sdk/sdk_flutter.dart';

MCP server referenced without pinned version: 'npx expo'.

Medium
Category
MCP Rug Pull
Confidence
70% confidence
Finding

npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.

Content

No source excerpt is available for this finding.

MCP server referenced without pinned version: 'npx expo'.

Medium
Category
MCP Rug Pull
Confidence
70% confidence
Finding

npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 708)May include surrounding context.

Response 201: monitoring_token, expires_at, allowed_event_categories, max_uses, bind_to_origin.

typescript
await fetch("https://verification.didit.me/v1/events/", {
  method: "POST",
  headers: { "X-Monitoring-Token": monitoringToken, "Content-Type": "application/json" },
  body: JSON.stringify({

External Transmission

Medium
Category
Data Exfiltration
Confidence
80% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · scripts/mint_sdk_tokens.py (reported line 57)May include surrounding context.

python
def post(path: str, payload: dict, api_key: str) -> dict:
    response = requests.post(
        f"{BASE_URL}{path}",
        headers={"x-api-key": api_key, "Content-Type": "application/json"},
        json=payload,