tuzi-update-claude-md
Pass
Audited by Gen Agent Trust Hub on Mar 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local bash commands including git log and git diff to analyze recent development activity, and wc to monitor documentation file length.
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its core workflow of summarizing repository history into behavioral guidelines.
- Ingestion points: Data is ingested from git diff outputs, git commit messages (logs), and the $ARGUMENTS variable.
- Boundary markers: No boundary markers or 'ignore' instructions are used to separate untrusted git content from the rule extraction logic.
- Capability inventory: The skill has the capability to modify CLAUDE.md and docs/CODING_RULES.md, which serve as instruction sets for future agent behavior, and can execute various git commands.
- Sanitization: No sanitization, escaping, or validation logic is applied to the content retrieved from the repository before it is processed and stored.
Audit Metadata