The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from GitHub Action logs and workflow files to identify CI inefficiencies. While this is the intended functionality, malicious content within these external sources could theoretically attempt to influence the agent's analysis or optimization suggestions.
Ingestion points: The script scripts/ci_audit.py fetches workflow definitions, and the gh api is used to retrieve job logs and organization billing data.
Boundary markers: No specific delimiters or safety instructions are provided to the agent to treat processed log content as untrusted data.
Capability inventory: The agent has the ability to write files, perform git operations (commit, push, merge), and execute GitHub CLI commands via the provided helper script.
Sanitization: The skill does not explicitly sanitize or validate the content of the logs before the agent processes them.
[COMMAND_EXECUTION]: The helper script scripts/ci_audit.py uses the subprocess module to execute the GitHub CLI (gh) for data collection. These commands are structured and targeted at specific GitHub API endpoints necessary for the CI audit functionality.

ci-audit