scope
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to analyze the contents of the local repository, including code, documentation, and tests, which constitutes an ingestion surface for untrusted data.
- Ingestion points: The section '1. Ground the request' in SKILL.md directs the agent to read repository instructions, docs, and code.
- Boundary markers: The instructions do not define clear delimiters or provide warnings to ignore potential instructions embedded within the repository files being read.
- Capability inventory: The agent has the capability to read all files in the repository and write output to a new or existing file (docs/scope.md).
- Sanitization: No sanitization, validation, or filtering of the repository content is required before the agent processes the information to generate requirements.
Audit Metadata