fec-alchemy
Pass
Audited by Gen Agent Trust Hub on Jun 24, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to run project-specific commands such as type checks, lints, tests, and builds to verify implemented changes (SKILL.md, Step 7).
- [EXTERNAL_DOWNLOADS]: The workflow involves analyzing external repositories, packages, and documentation to serve as reference inspiration for project improvements (SKILL.md, Step 1).
- [PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by processing external, untrusted 'reference systems'. This is mitigated by explicit instructions to prioritize the local repository as the source of truth and to perform original redesign rather than direct copying (SKILL.md, Step 1 and Step 6).
Audit Metadata