payment-integration
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [Indirect Prompt Injection] (LOW): The skill is designed to handle untrusted data from payment provider webhooks.\n
- Ingestion points: Webhook event handlers described in
references/sepay/webhooks.mdandreferences/polar/webhooks.md.\n - Boundary markers: The implementation workflow mandates the use of verification scripts (
scripts/sepay-webhook-verify.jsandscripts/polar-webhook-verify.js) to validate payloads.\n - Capability inventory: The skill handles payments and subscriptions, which involves network API calls and script execution.\n
- Sanitization: Explicit mention of signature and authenticity verification for both SePay and Polar platforms is provided as a mandatory step.\n- [Prompt Injection] (SAFE): No evidence of instructions designed to bypass agent safety protocols or override core behavior was found.\n- [Data Exposure & Exfiltration] (SAFE): The documentation mentions API keys and OAuth2 in a standard context without hardcoding sensitive credentials or targeting unauthorized file paths.\n- [Unverifiable Dependencies & Remote Code Execution] (SAFE): While helper scripts and SDKs are referenced, there are no patterns indicating the download or execution of untrusted remote code.
Audit Metadata