What problem does it solve?
It reduces execution risk when you already have a written plan by ensuring you review it first, run tasks in controlled batches, and produce clear checkpoint reports for architect or partner feedback.
Core Features & Use Cases
- Plan-first critical review: Reads a provided plan, flags questions or concerns, and asks your human partner before starting when gaps exist.
- Batch execution with verification: Executes the first N tasks (default first 3), follows the plan’s bite-sized steps precisely, and runs verifications specified in the plan.
- Checkpoint reporting loop: Reports implemented work and verification output, then waits for feedback before continuing with the next batch.
- Guardrails for stopping early: Stops immediately when dependencies are missing, tests fail, instructions are unclear, or verification repeatedly fails.
- Development completion handoff: After all tasks complete and verify, transitions to the required sub-skill superpowers:finishing-a-development-branch for finalization.
Quick Start
Use the executing-plans skill when you have a written implementation plan and want the AI to execute the plan in batches with verification checkpoints and a readiness-for-feedback report between batches.
