The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill relies on external Node.js packages, specifically ruv-swarm and claude-flow, which are fetched and executed at runtime using npx or global installation.
[COMMAND_EXECUTION]: The skill makes extensive use of the GitHub CLI (gh) and shell commands to perform repository operations, such as creating issues, commenting on pull requests, and downloading workflow logs.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from multiple external sources, creating a vulnerability surface where embedded instructions could influence the agent's behavior.
Ingestion points: Ingests pull request data through gh pr view, workflow logs via gh run view, and commits through codebase analysis commands.
Boundary markers: No explicit boundary markers or instructions to ignore embedded commands were identified in the documented CLI interactions.
Capability inventory: The skill possesses capabilities to write to the repository (creating PRs and code fixes), create GitHub issues, and post comments.
Sanitization: Data is parsed using jq but then interpolated directly into gh command arguments (e.g., gh issue create --title "$(_jq '.title')").
[DYNAMIC_EXECUTION]: The skill utilizes dynamic execution patterns, including generating GitHub Action test matrices at runtime and suggesting 'self-healing' code modifications based on CI/CD failure analysis.

github-workflow-automation