The Agent Skills Directory

[COMMAND_EXECUTION]: The documentation provides examples for executing shell commands using the bash tool, including curl for HTTP requests, git for repository management, and npm for project builds. These are documented as standard development operations within the agentic workflow context.
[DYNAMIC_EXECUTION]: The skill illustrates how to develop and integrate custom Model Context Protocol (MCP) servers using Node.js. It responsibly includes a dedicated 'Security Considerations' section that provides code examples for input validation, path sanitization (to prevent traversal), and audit logging.
[INDIRECT_PROMPT_INJECTION]: The documented tools for web searching, browser automation (Playwright), and filesystem access define an attack surface where an agent could ingest untrusted data containing malicious instructions. The skill acknowledges this risk by documenting mitigation patterns such as input validation and boundary checks.
[SAFE]: All external references and repository links target well-known services (such as official GitHub documentation domains) or resources belonging to the skill author 'hack23' (e.g., the riksdagsmonitor repository). No suspicious or typosquatted domains were identified.

GitHub Agentic Workflows Tools Ecosystem