The Agent Skills Directory

[COMMAND_EXECUTION]: The skill frequently executes shell commands using the GitHub CLI (gh), jq, and python3 to automate repository analysis and configuration updates.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes untrusted data from external repositories.
Ingestion points: Reads repository names, pull request titles, branch names, and GitHub Actions workflow file contents using gh repo list, gh pr list, and gh api calls in SKILL.md.
Boundary markers: The skill does not employ delimiters or specific instructions to the agent to ignore potential commands embedded within the fetched GitHub data.
Capability inventory: The skill possesses write capabilities, including the ability to enable security alerts and overwrite workflow files via gh api --method PUT (e.g., in the semgrep-syntax and dependabot-config fix steps).
Sanitization: There is no evidence of sanitization or escaping of the data fetched from GitHub before it is included in the output report or used in fix logic.

repo-audit