The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes standard git commands (git diff, git log, git remote) and GitHub CLI (gh repo view) to determine the base branch and change context. These are routine operations for developer tools.
[COMMAND_EXECUTION]: The skill performs project discovery by running make help and searching for common configuration files (pyproject.toml, package.json, etc.) to provide contextually relevant testing instructions. This is legitimate behavior for a test-plan generator.
[DATA_EXFILTRATION]: No network operations to non-whitelisted or suspicious domains were detected. Network activity is limited to standard git and gh interactions with the origin repository.
[CREDENTIALS_UNSAFE]: No hardcoded secrets, API keys, or attempts to access sensitive credential files (like ~/.ssh or .env) were found. The skill instructs users to provide environment variables inline for specific commands if necessary, which is a common manual testing pattern.
[PROMPT_INJECTION]: No attempts to override system prompts or bypass safety filters were detected. The instructions follow a logical workflow for generating documentation.
[NO_CODE]: The skill does not bundle any external scripts or executables, relying entirely on shell commands provided as instructions for the agent to execute in the user's environment.

cmd-pr-test-plan