draft-technical
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted project data including documentation notes, principles, and source code, which serves as an injection surface.\n
- Ingestion points: Reads files from docs//notes/*, EXPERIENCE.final.md, project documentation (PRINCIPLES.md, AGENTS.md, CONTRIBUTING.md), and the project's source code (File: SKILL.md).\n
- Boundary markers: The skill does not define specific delimiters or instructions to ignore embedded prompts within the files it reads.\n
- Capability inventory: The agent can execute git commands (fetch, log) and write files to the docs/ directory.\n
- Sanitization: No explicit sanitization or filtering of the ingested file content is mentioned.\n- [COMMAND_EXECUTION]: The skill instructs the agent to use the git CLI to perform operations like fetching updates and checking logs.\n
- Evidence: Instructs the use of git fetch and git log --oneline to synchronize with the main integration branch before designing (File: SKILL.md).
Audit Metadata