executing-plans
Use Cautionnpx machina-cli add skill Ibrahim-3d/conductor-orchestrator-superpowers/executing-plans --openclawExecuting Plans
Overview
Load plan, review critically, execute tasks in batches, report for review between batches.
Core principle: Batch execution with checkpoints for architect review.
Announce at start: "I'm using the executing-plans skill to implement this plan."
The Process
Step 1: Load and Review Plan
- Read plan file
- Review critically - identify any questions or concerns about the plan
- If concerns: Raise them with your human partner before starting
- If no concerns: Create TodoWrite and proceed
Step 2: Execute Batch
Default: First 3 tasks
For each task:
- Mark as in_progress
- Follow each step exactly (plan has bite-sized steps)
- Run verifications as specified
- Mark as completed
Step 3: Report
When batch complete:
- Show what was implemented
- Show verification output
- Say: "Ready for feedback."
Step 4: Continue
Based on feedback:
- Apply changes if needed
- Execute next batch
- Repeat until complete
Step 5: Complete Development
After all tasks complete and verified:
- Announce: "I'm using the finishing-a-development-branch skill to complete this work."
- REQUIRED SUB-SKILL: Use superpowers:finishing-a-development-branch
- Follow that skill to verify tests, present options, execute choice
When to Stop and Ask for Help
STOP executing immediately when:
- Hit a blocker mid-batch (missing dependency, test fails, instruction unclear)
- Plan has critical gaps preventing starting
- You don't understand an instruction
- Verification fails repeatedly
Ask for clarification rather than guessing.
When to Revisit Earlier Steps
Return to Review (Step 1) when:
- Partner updates the plan based on your feedback
- Fundamental approach needs rethinking
Don't force through blockers - stop and ask.
Remember
- Review plan critically first
- Follow plan steps exactly
- Don't skip verifications
- Reference skills when plan says to
- Between batches: just report and wait
- Stop when blocked, don't guess
- Never start implementation on main/master branch without explicit user consent
Conductor Integration (Autonomous Mode)
When invoked with --plan, --track-dir, and --metadata parameters (from Conductor orchestrator):
- Read plan from
--planpath - Execute ALL tasks (not batches of 3) — run autonomously
- Do NOT stop for human feedback between batches
- After each task: use Edit tool to mark
[x]in plan.md with commit SHA - After all tasks: update
--metadatacheckpoint toEXECUTE: PASSED - Return concise verdict:
{"verdict": "PASS", "tasks_completed": N} - If
--resume-fromis provided, skip tasks before that task ID
When these parameters are absent, fall back to the standalone batch workflow above.
Integration
Required workflow skills:
- superpowers:using-git-worktrees - REQUIRED: Set up isolated workspace before starting
- superpowers:writing-plans - Creates the plan this skill executes
- superpowers:finishing-a-development-branch - Complete development after all tasks
Source
git clone https://github.com/Ibrahim-3d/conductor-orchestrator-superpowers/blob/master/skills/executing-plans/SKILL.mdView on GitHub Overview
Executing-plans loads a written plan, reviews it critically, and executes tasks in controlled batches with review checkpoints. It emphasizes batch execution, verification between batches, and clear communication to keep stakeholders aligned.
How This Skill Works
The skill loads the plan, performs a critical review, and, if no concerns exist, creates a Todo and proceeds. It then runs tasks in batches (default first 3), marks progress, runs verifications, and reports between batches before continuing. After all tasks are verified, it announces completion and uses the finishing-a-development-branch sub-skill to finalize.
When to Use It
- You have a written implementation plan and want to execute it in a separate session with review checkpoints.
- You need controlled batch execution (first 3 tasks) with verifications between batches.
- You need to raise questions or concerns about the plan before starting execution.
- You are planning to complete development after all tasks are finished using a dedicated finalization step.
- You want autonomous execution via Conductor Integration to run all tasks or handle plan updates.
Quick Start
- Step 1: Load the plan file, review critically, and raise concerns if any.
- Step 2: Start execution with the first batch (default 3 tasks), marking in_progress and verifications for each.
- Step 3: After batch completion, report what was implemented and verification results, then proceed to the next batch or finish.
Best Practices
- Load and review the plan critically before starting.
- Raise any concerns or questions to the human partner prior to execution.
- Follow each bite-sized step exactly as written in the plan.
- Run verifications after each task and do not skip verifications.
- Announce you are starting and, after completion, invoke finishing-a-development-branch for finalization.
Example Use Cases
- Implementing a new feature by executing a 12-task plan in 3 batches and validating each batch with tests.
- Identifying a missing dependency during Step 1 and halting the plan to request clarification before proceeding.
- Reporting completed tasks and verification outputs to a partner and awaiting feedback.
- Progressing through multiple batches and using the finishing-a-development-branch sub-skill at the end to finalize.
- Using Conductor Autonomous Mode with --plan to execute all tasks without intermediate human feedback, updating plan.md with commit SHAs after each task.
