The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses git revert and git log to manage project state and documentation synchronization. This capability is strictly gated by a multi-stage approval process (Phase 1 and Phase 3), requiring the user to explicitly confirm target selection and the specific list of commit SHAs before any execution occurs.\n- [PROMPT_INJECTION]: The skill reads data from local project files like plan.md and tracks.md to resolve commit SHAs and work item statuses. This introduces a surface for indirect prompt injection where a malicious file could attempt to influence the agent's behavior. However, the skill mitigates this by validating extracted SHAs against the Git history and presenting the final execution plan to the user for verification.

conductor-revert