The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the GitHub CLI (gh) to interact with repositories for auditing and metadata updates. This is the primary function of the skill and aligns with its stated purpose of managing GitHub project quality.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from local files and GitHub repositories (e.g., READMEs, issue templates, and metadata) to perform its audit functions. While this creates a vulnerability surface where malicious repo content could influence the agent, the behavior is intrinsic to the skill's utility and includes clear workflow phases to limit unintended actions.
Ingestion points: Reads repository files and metadata via local access or gh CLI commands.
Boundary markers: None explicitly defined in the instructions, though the audit follows a structured scoring system.
Capability inventory: Includes file system write access and repository configuration changes via the gh CLI.
Sanitization: Not explicitly documented; the skill relies on the agent's ability to analyze and improve existing prose.
[EXTERNAL_DOWNLOADS]: The skill references and invokes other skills (metadata-check and readability-check) authored by the same developer. These are legitimate dependencies within the skill ecosystem for providing specialized quality checks.

github-repo