release
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes commit messages via git logs to generate a structured changelog. This creates a surface where malicious instructions embedded in commit messages could influence the agent's summary or the content of the resulting pull request body.\n
- Ingestion points: The agent reads commit history using
git log origin/main..dev --onelinein SKILL.md.\n - Boundary markers: No delimiters or specific instructions to ignore embedded commands within the log data are present.\n
- Capability inventory: The skill has the ability to modify local files (
package.json,server.json), push code to remote repositories, and create pull requests.\n - Sanitization: There is no evidence of filtering or sanitizing the commit message content before it is interpolated into the changelog generation prompt.\n- [COMMAND_EXECUTION]: The skill executes several shell commands to manage the release workflow, including git operations and project builds.\n
- Evidence: Uses
npm install --package-lock-onlyandnpm run buildto manage dependencies and verify project integrity.\n - Evidence: Performs branch management and code synchronization using
git fetch,git checkout, andgit push.
Audit Metadata