plan-review
Pass
Audited by Gen Agent Trust Hub on Nov 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted project data (task records, code, and rules) which could contain embedded instructions intended to influence the agent's behavior.
- Ingestion points: The skill reads task records, project rules, current code evidence, and qualified ticket IDs from the local environment.
- Boundary markers: The instructions do not specify any delimiters or safety markers to differentiate between data to be reviewed and instructions to be followed.
- Capability inventory: The agent can read project files and use research and design tools to investigate the codebase.
- Sanitization: There are no requirements or instructions for sanitizing or escaping the data read from external project files.
Audit Metadata