AgentSkillsCN

executing-plans

在拥有书面的实施计划并按计划推进，且每项任务之间设有评审节点时使用。

SKILL.md
--- frontmatter
name: executing-plans
description: Use when you have a written implementation plan to execute, with review checkpoints between tasks

Executing Plans

Overview

Load plan, review critically, execute tasks in batches, report for review between batches.

Core principle: Batch execution with checkpoints. Review after every batch.

Steps

Step 1: Load and Review Plan

  1. •Read plan file completely
  2. •Review critically — identify questions or concerns
  3. •If concerns: raise with the user before starting
  4. •If clear: proceed

Step 2: Execute Batch (3 tasks default)

For each task:

  1. •Mark as in_progress
  2. •Follow each step exactly (plan has bite-sized steps)
  3. •Run verifications as specified
  4. •Mark as completed

Step 3: Two-Stage Review

After each task (or batch):

Stage 1 — Spec Compliance:

  • •Does implementation match plan requirements?
  • •Any missing or extra functionality?
  • •If issues found: fix before moving to Stage 2

Stage 2 — Code Quality:

  • •Clean architecture?
  • •Magic numbers, naming, duplication?
  • •If issues found: fix before moving to next task

Step 4: Report and Continue

When batch complete:

  • •Show what was implemented
  • •Show verification output
  • •Say: "Ready for feedback."
  • •Apply feedback, execute next batch
  • •Repeat until complete

Step 5: Complete Development

After all tasks verified:

  • •Use finishing-work skill to handle merge/PR/cleanup

When to STOP

Stop executing immediately when:

  • •Hit a blocker (missing dependency, test fails, unclear instruction)
  • •Plan has critical gaps
  • •You don't understand an instruction
  • •Verification fails repeatedly

Ask for clarification rather than guessing.

Subagent Execution (Optional)

For independent tasks, dispatch subagents:

  • •Provide full task text (don't make subagent read plan file)
  • •Include scene-setting context
  • •Run spec compliance review after each subagent completes
  • •Run code quality review after spec compliance passes
  • •Answer subagent questions before letting them proceed

Red Flags

  • •Skipping verifications
  • •Guessing when blocked instead of asking
  • •Starting on main/master without user consent
  • •Batching without review checkpoints
  • •Proceeding with unfixed review issues