The Agent Skills Directory

[SAFE]: The skill uses official GitHub CLI (gh) and REST APIs for its operations, targeting well-known documentation and official command-line tools.
[COMMAND_EXECUTION]: The skill is designed to use the gh CLI to interact with GitHub services. These operations are constrained by the skill's defined boundaries to release, tag, and asset management.
[INDIRECT_PROMPT_INJECTION]: The skill has a potential surface for indirect prompt injection as it processes untrusted data from merged Pull Requests to generate changelogs and release notes.
Ingestion points: Merged Pull Request bodies and titles fetched via the GitHub API in SKILL.md.
Boundary markers: None explicitly defined in the instructions to separate processed data from agent instructions.
Capability inventory: Uses gh CLI for release creation, editing, and asset uploading as described in SKILL.md.
Sanitization: No specific sanitization or filtering of PR content is mentioned before interpolation into release notes. This is a standard risk factor for automated changelog generators and is considered low risk in this context.

release-manager