The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from Git commit messages and GitHub pull request bodies to assist in generating release notes. This creates a surface for indirect prompt injection where malicious content in a repository's history could attempt to influence the agent's output.\n
Ingestion points: Data is collected via git log, gh api for commit-associated pull requests, and gh pr view for pull request titles and bodies.\n
Boundary markers: Technical delimiters are absent; the skill relies on natural language instructions to prioritize the code diff as the source of truth.\n
Capability inventory: The agent can modify remote content via gh release edit, perform network operations via git fetch and the GitHub API, and access the local file system using mktemp.\n
Sanitization: No explicit sanitization or filtering of the external text is performed before processing.\n- [COMMAND_EXECUTION]: The skill executes various Git and GitHub CLI (gh) commands to analyze the repository and update release data. It follows best practices by quoting variables and using the -- separator to mitigate command injection risks.\n- [EXTERNAL_DOWNLOADS]: The skill performs network operations to fetch repository metadata and commit history from GitHub, which is a well-known service.

release-notes