The Agent Skills Directory

[COMMAND_EXECUTION]: The skill constructs and executes shell commands using the GitHub CLI (gh) and git. It uses heredocs and subshells to generate pull request content and verify repository status. It also utilizes command chaining with xargs to perform bulk permission checks across organization teams.
[EXTERNAL_DOWNLOADS]: The skill references and provides installation instructions for external dependencies from the soia-team repository. These include soia-dev-review-panel and soia-dev-fix-loop, which are integrated into the PR review and fix workflows.
[PROMPT_INJECTION]: The skill has a surface for indirect prompt injection because it ingests untrusted data from GitHub that is then used to guide automated actions.
Ingestion points: The skill fetches pull request bodies, inline diff comments, and conversation-tab comments via gh api and gh pr view in SKILL.md.
Boundary markers: Absent. There are no instructions for the agent to wrap external data in delimiters or to ignore potential instructions embedded within the comments.
Capability inventory: The skill has high-impact capabilities including git push, gh pr merge, and the ability to modify repository collaborator permissions via gh api.
Sanitization: Absent. The skill passes fetched comments directly to the soia-dev-fix-loop dependency without explicit filtering or validation of the content.

soia-dev-github-ops