The Agent Skills Directory

[SAFE]: No malicious patterns or security vulnerabilities were detected. The skill is designed with a strong focus on authorization and user oversight.
[COMMAND_EXECUTION]: The skill executes standard git and GitHub CLI commands as intended. It includes specific logic to prevent the use of high-risk flags (like --force) and requires additional confirmation for protected branches.
[PROMPT_INJECTION]: The instructions reinforce security boundaries by designating this skill as the sole authorized path for push operations and explicitly forbidding the agent from attempting to bypass these restrictions.

push-ci