git-workflow-and-versioning
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs agents to analyze code changes and provide structure summaries of work done. This creates a surface where malicious instructions embedded in file diffs or code comments could influence the agent's behavior.
- Ingestion points: Git diffs via
git diff --stagedand overall file content analysis during the commit/PR process. - Boundary markers: The skill does not provide specific delimiters or instructions for the agent to distinguish between its own operational instructions and potentially malicious content within the code being summarized.
- Capability inventory: The agent has access to
git,npm, andnpxtools for testing and verification. - Sanitization: No sanitization or filtering logic is prescribed for processing untrusted code data before interpolation into summaries.
- [SAFE]: The skill actively encourages security best practices by instructing agents to audit for hardcoded secrets (e.g.,
password,api_key,token) before committing and providing a robust.gitignoretemplate to exclude sensitive environment files and private keys.
Audit Metadata