The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes a Python script scripts/check_pages_setup.py that executes local git commands to verify the repository configuration.
Evidence: The script uses subprocess.run to call git remote get-url, git rev-parse, git ls-remote, and git ls-tree. These commands are executed without shell=True and are used exclusively for diagnostic purposes to ensure the repository is correctly configured for GitHub Pages deployment.
[INDIRECT_PROMPT_INJECTION]: The skill processes external repository data (workflows and configuration files), which constitutes a potential surface for indirect prompt injection.
Ingestion points: scripts/check_pages_setup.py reads content from GitHub Actions workflow files and static site generator configuration files (such as mkdocs.yml, docusaurus.config.js, etc.).
Capability inventory: The script's capabilities are limited to reading local files and executing local git commands for diagnostic analysis.
Boundary markers: No explicit boundary markers are present in the script's reading logic.
Sanitization: The script employs specific regular expressions to extract configuration values, which inherently restricts the data processed to expected patterns and mitigates the risk of processing arbitrary malicious instructions.

docs-site