github-mcp

Pass

Audited by Gen Agent Trust Hub on Jun 22, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the setup of the GitHub MCP server using npx with the @modelcontextprotocol/server-github package or via Docker from ghcr.io/modelcontextprotocol/server-github. These are well-known and official sources for the Model Context Protocol ecosystem.- [PROMPT_INJECTION]: The skill defines tools that ingest untrusted data from GitHub (e.g., issue bodies, code search results, pull request files), creating an indirect prompt injection surface. This is a characteristic of the skill's intended functionality for repository management.
  • Ingestion points: list_issues, search_code, get_pull_request_files (found in SKILL.md)
  • Boundary markers: None identified in the documentation.
  • Capability inventory: Substantial write access including merging PRs and triggering workflows (SKILL.md).
  • Sanitization: No explicit sanitization steps are defined in the instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 22, 2026, 07:46 AM
Security Audit — agent-trust-hub — github-mcp