The Agent Skills Directory

[External Downloads] (LOW): The skill configures the GitHub MCP server using npx @modelcontextprotocol/server-github. While npx fetches and executes remote packages, this is the standard and recommended deployment method for the official Model Context Protocol GitHub server.
[Indirect Prompt Injection] (LOW): The skill is designed to ingest and process untrusted data from GitHub repositories (such as pull request bodies, comments, and issue descriptions) and possesses write capabilities.
Ingestion points: Tools like mcp__github__get_pull_request, mcp__github__get_issue, and mcp__github__get_pull_request_comments bring external text into the agent context.
Boundary markers: The instructions do not define explicit delimiters or warnings to ignore instructions embedded within the GitHub data.
Capability inventory: The skill has the ability to write data back to GitHub using mcp__github__issue_write and mcp__github__create_pull_request_review.
Sanitization: No explicit sanitization or validation of the retrieved content is described.
[Data Exposure & Exfiltration] (SAFE): No hardcoded credentials or sensitive local file paths were detected. The skill correctly instructs the user to use environment variables (GITHUB_TOKEN) for authentication.

github-integration