The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses uv run to execute a local Python script scripts/mcp/github_search.py that handles the logic for searching GitHub content. This is the intended behavior for the tool.- [EXTERNAL_DOWNLOADS]: The skill communicates with GitHub's official API to fetch data. As GitHub is a well-known service and the downloads are limited to search results, this is documented as safe.- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data (search results from GitHub) which represents an attack surface. Ingestion points: Search results from GitHub code, issues, and PRs enter the agent's context through scripts/mcp/github_search.py. Boundary markers: None identified in the skill instructions. Capability inventory: The agent has access to the Bash tool and can execute commands via uv run. Sanitization: No specific sanitization or filtering of the search results is mentioned in the skill documentation.

github-search