github-mcp
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill documentation in
SKILL.mdspecifies the use of the official GitHub MCP server image (ghcr.io/github/github-mcp-server). This is a reference to a well-known and trusted service for GitHub operations. - [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of data from external sources (GitHub repositories, issues, and pull requests) which are controlled by third parties and could contain malicious instructions.
- Ingestion points: Untrusted data enters the agent context through tools like
issue_read,pull_request_read, andsearch_codeas described inSKILL.md. - Boundary markers: The skill files do not define explicit delimiters or 'ignore instructions' warnings to wrap content fetched from the GitHub API.
- Capability inventory: The skill possesses significant capabilities including file system writes (
create_or_update_file,push_files), branch management, and arbitrary command execution via theBashtool. - Sanitization: There is no explicit sanitization logic in the provided scripts to filter or escape instructions embedded in the retrieved GitHub data.
Audit Metadata