github-repo-management
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses local shell commands to perform git operations including
git init,git add,git commit,git push, andgit branch. These are standard developer tools and their use is scoped to repository management tasks. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sources when listing Pull Requests or fetching repository details via GitHub MCP.
- Ingestion points: GitHub MCP tools (list PRs, get repo info) and
git remote get-url. - Boundary markers: Absent.
- Capability inventory: Local git execution (
commit,push,branch -d) and GitHub MCP write actions (create_repository,merge). - Sanitization: Absent, although the skill instructs the agent to "summarize" the data, providing a natural buffer. This risk is inherent to the primary purpose of the skill.
- [SAFE]: The skill includes explicit security constraints, such as prohibiting the output or logging of tokens and secrets. It also mandates user confirmation for high-risk destructive actions like
git push --forceor deleting branches and repositories.
Audit Metadata