payment-status-sync
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from Stripe webhooks and API responses, which constitutes an indirect prompt injection surface.
- Ingestion points: Stripe webhook event objects in patterns 1 and 2, and checkout session IDs from URL parameters in pattern 5, both within SKILL.md.
- Boundary markers: Absent; the instructions do not explicitly mandate delimiters or warnings for the agent when processing this external data.
- Capability inventory: The skill performs database write operations (db.subscription.update) and network requests to the Stripe API throughout SKILL.md.
- Sanitization: Basic TypeScript type casting is used, but specific sanitization of external metadata is not explicitly detailed, although the data is used primarily for status and timestamp updates.
Audit Metadata