The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/repo_audit.py executes git and gh commands via subprocess.run. These calls are limited to read-only operations, such as listing tags, checking remote URLs, and performing GitHub API GET requests. The commands use argument lists and avoid shell execution, preventing shell injection vulnerabilities.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the repository's files (e.g., workflow definitions, CODEOWNERS, and license files) to perform its checks. Although this provides an attack surface for indirect prompt injection, the risk is mitigated by the script's read-only nature and its use of structured parsing (JSON/Regex) rather than direct interpolation into agent instructions.

repo-audit