The Agent Skills Directory

[COMMAND_EXECUTION]: The skill invokes the gh (GitHub CLI) and git command-line tools via execFileSync to fetch repository information, perform GraphQL queries, and manage local project state. These actions are limited to the repository scope and are necessary for the skill's primary function.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sources including GitHub issue bodies, pull request descriptions, and repository configuration files like issue templates.
Ingestion points: Data is pulled from the GitHub API (github.ts) and read from repository-maintained files like CONTRIBUTING.md and .github/ISSUE_TEMPLATE/ (derive.ts).
Boundary markers: The skill generates structured reports using markdown headers and manages persistent notes with YAML frontmatter to separate contributor input from maintainer analysis.
Capability inventory: The skill possesses the capability to write files to the project directory and execute Git/GitHub CLI operations across the workspace.
Sanitization: The implementation includes sanitization logic for phrases derived from templates (derive.ts) and uses localized sentiment analysis to evaluate contributor input without direct execution of external content.

open-source-maintainer