executing-plans
Batch-based execution workflow that enforces incremental progress and constant verification.
Install
mkdir -p .claude/skills/executing-plans-liujiarui0918 && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/11365" && unzip -o skill.zip -d .claude/skills/executing-plans-liujiarui0918 && rm skill.zipInstalls to .claude/skills/executing-plans-liujiarui0918
Activation
This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.
Use to execute a written plan. Run tasks in batches with human checkpoints.Key capabilities
- →Execute tasks in batches
- →Verify task success
- →Manage human checkpoints
- →Report progress
How it works
It runs tasks in batches of 1-3, followed by verification and human feedback.
Inputs & outputs
When to use executing-plans
- →Executing coded plans
- →Incremental task delivery
- →Verifying code changes
About this skill
Executing Plans
Iron Law: 一次 1-3 个任务,verify,汇报,等用户确认,再下一批。不要一口气干完。
触发时机
- 拿到一份 [[writing-plans]] 产出的 plan
- 用户说"按 plan 执行"
- 你已经有清晰任务列表 + 验证标准
不触发
- 还在 brainstorming → [[brainstorming]]
- plan 都没写完 → [[writing-plans]]
- 任务彼此独立 ≥2 个,且都可并行 → [[dispatching-parallel-agents]]
- 任务复杂需要 subagent 隔离 → [[subagent-driven-development]]
流程
取 1-3 个任务 (batch)
↓
逐个执行:Edit/Write → 跑 Verify 命令 → 检查 Success criterion
↓
汇报:列出做了什么、verify 输出、是否符合预期
↓
等用户 ack
↓
下一批
Batch 选择规则
- 首批:选 1 个最小、最独立的任务作为热身,验证理解一致
- 中段:每批 2-3 个,相关或顺序紧密的放一批
- 末批:剩余清扫 + 端到端验证
Verify 必须真跑
- 不要"看起来应该对了"——跑命令
- Verify 命令在 plan 里写好了,照着跑
- 如果 plan 没写 verify → 回到 [[writing-plans]] 补上,不要自己脑补
汇报格式
## Batch N
- Task A:done. Verify: `<command>` → `<output snippet>`. Success: ✓
- Task B:done. Verify: `<command>` → `<output snippet>`. Success: ✓
- Task C:done. Verify: `<command>` → `<output snippet>`. Success: ✓
继续?
Red Flags
- 连干 5 个任务没汇报 → 停,立即汇报
- 跳过 verify 直接下一题 → 停,回去补 verify
- verify 失败但你"差不多就行了" → 停,问用户
- 自己改 plan 没告诉用户 → 停,先说
失败处理
任务失败时:
- 立刻停,不要继续下一题
- 报告:失败的任务、verify 输出、你的假设
- 问用户:「修这里?跳过?改 plan?」
- 不要自作主张回滚或重做
与 subagent 的边界
主线程跑 [[executing-plans]] 时,你就是执行者。如果某个 task 涉及大量探索 / 隔离 context / 多任务并行,升级到:
- [[subagent-driven-development]]:每任务一个干净 subagent + review
- [[dispatching-parallel-agents]]:并发派 ≥2 个独立任务
MUST
- 1-3 个任务一批
- 每个任务都跑 verify
- 失败立刻停,问人
- 汇报格式固定,便于扫读
MUST NOT
- 不要连开 5 个 Edit 不停顿
- 不要假设 verify 通过
- 不要不告而别地改 plan
- 不要把 verify 失败"绕过去"
When not to use it
- →Brainstorming
- →Parallel independent tasks
Prerequisites
Limitations
- →1-3 tasks per batch
- →Requires human ack
How it compares
It enforces strict human checkpoints to prevent over-execution or drift.
Compared to similar skills
executing-plans side by side with the closest alternatives in the catalog.
| Skill | Installs | Updated | Safety | Difficulty |
|---|---|---|---|---|
| executing-plans (this skill) | 0 | 2mo | No flags | Beginner |
| github-script | 1 | 3mo | No flags | Intermediate |
| spec-to-done | 0 | 1mo | No flags | Advanced |
| lfg | 0 | 3mo | No flags | Advanced |
Try saying
Example prompts that trigger this skill in your AI assistant.
You might also like
github-script
githubnext
Best practices for writing JavaScript code for GitHub Actions using github-script
spec-to-done
monkey1sai
Use when a brainstormed spec already exists under docs/superpowers/specs/ and the user asks to autonomously drive it to a merged PR (e.g. 「跑 spec-to-done」「用 spec-to-done 跑 <spec 路徑>」), or when resuming a previously held spec-to-done run.
lfg
nabilban
Full autonomous engineering workflow
executing-plans
obra
Use when you have a written implementation plan to execute in a separate session with review checkpoints
schedules
windmill-labs
MUST use when configuring schedules.
task-think
telegramdesktop
Orchestrate a multi-phase implementation workflow for this repository with artifact files under .ai/task-slug and optional fresh codex exec child runs per phase. Use when the user wants one prompt to drive context gathering, planning, implementation, verification, and review iterations while keeping the main session context clean.