read-github
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill fetches and processes external content from any GitHub repository or URL. * Ingestion points: Content is ingested from gitmcp.io and external URLs via scripts/gitmcp.py. * Boundary markers: The skill lacks instructions for the agent to treat fetched content as untrusted data or use delimiters to prevent instruction override. * Capability inventory: The skill relies on local script execution and network access. * Sanitization: No evidence of content sanitization is provided.
- [COMMAND_EXECUTION]: The skill instructions direct the agent to run a local script (scripts/gitmcp.py) using shell commands, incorporating potentially untrusted repository paths and search queries.
- [DATA_EXFILTRATION]: The skill performs network operations to a non-whitelisted domain (gitmcp.io) to fetch repository data and allows fetching content from arbitrary external URLs.
Audit Metadata