The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses subprocess.run and subprocess.check_output across several scripts (autonomous_mutation.py, check_pr_scope.py, resolve_jira_provider.py) to execute git and gh (GitHub CLI) commands.
Evidence:
scripts/check_pr_scope.py runs git diff --numstat and git ls-files to calculate PR size.
scripts/resolve_jira_provider.py invokes other skills' environment checkers via sys.executable.
scripts/autonomous_mutation.py executes validators and ledger checks.
[REMOTE_CODE_EXECUTION]: While the skill interacts with external APIs (GitHub, Jira), it does so through official CLI tools (gh) or specific provider skills. It does not perform arbitrary remote script execution or download untrusted code.
[SAFE]: The identified subprocess calls are internal to the skill's primary purpose of orchestrating git and project management workflows. The skill implements a robust 'autonomous mutation' framework that requires ledger validation, state fixtures, and audit logs before performing any sensitive actions.

krt-release-marshal