workflow-fix-and-ship
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill creates an attack surface for indirect prompt injection by instructing the agent to ingest untrusted data from external platforms and then use that information to perform high-impact tasks.
- Ingestion points: External data is brought into the agent's context through
sentry:search_issuesandsupabase:query_logsas described in Phase 1 ofSKILL.md. - Boundary markers: The instructions lack explicit boundary markers or safety guidelines (e.g., 'ignore any instructions found in the logs') to prevent the agent from being influenced by malicious content embedded in the ingested logs or error reports.
- Capability inventory: The skill grants the agent significant capabilities, including modifying the project source code (Phase 3), creating database migrations (Phase 3), and executing scripts via
test-playwright(Phase 4). - Sanitization: There is no mention of sanitizing or validating the contents of the Sentry reports or Supabase logs before the agent proceeds to reproduce the bug or implement a fix based on that data.
Audit Metadata