AgentStack
Browse Sign in
Browse Why AgentStack Sell Docs
Sign in
SKILL verified MIT Self-run

Plan Action

skill-donglinfei-debug-claude-plan-action-skill-skill-files · by donglinfei-debug

全盘任务规划 — 分析需求、拆解子任务、调度 Agent 和 Skill、输出执行方案

No reviews yet
0 installs
30 views
0.0% view→install

Install

$ agentstack add skill-donglinfei-debug-claude-plan-action-skill-skill-files

✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets No
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

View the full security report →

Verified badge

Passed review? Show it. Paste this badge into your README, it links to the public security report.

AgentStack Verified badge Links to your public security report.
[![AgentStack Verified](https://agentstack.voostack.com/badges/verified.svg)](https://agentstack.voostack.com/security/report/skill-donglinfei-debug-claude-plan-action-skill-skill-files)

Reliability & compatibility

Security review passed
0 installs to date
no reviews yet
3mo ago

Declared compatibility

Claude CodeClaude Desktop

Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.

Preview Execution monitoring

We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.

How agent discovery & health will work →
Are you the author of Plan Action? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

/plan-action — 全盘任务规划

当用户输入 /plan-action 时,进入结构化规划模式。规划阶段通过 Workflow 工具调度 pre-planning-orchestrator 实现,规划完成经你确认后,转入 controller-workflow 执行。

执行流程(两个 Workflow 阶段)

前置准备:收集审计数据

在调 Workflow 之前,先读取以下数据作为 Workflow 的输入参数:

  1. 读取 AGENT_REGISTRY.json → 得到 agentRegistry
  2. 扫描 skills/ 目录(含 CLAUDE.md 技能表)→ 得到 availableSkills
  3. 判断 projectType(新项目 / 已有项目变更)

第 1-4 步:调用 Workflow( pre-planning-orchestrator )

调用 Workflow 工具,执行前置规划编排:

Workflow({
  name: "pre-planning-orchestrator",
  args: {
    userRequest: "",
    projectType: "新项目 / 已有项目变更",
    availableSkills: [...],
    agentRegistry: { agents: [...] },
    techConstraints: "",
    hasCodebase: true/false
  }
})

Workflow 执行过程中将依次显示:

  • [L1] Controller: 需求解析与Agent分解中...
  • [L1] Controller: 技能审计中...
  • [L1] Controller: 可行性评估中...
  • [L1] Controller: 执行计划生成中...

Workflow 返回结构化结果后,按以下 5 模块格式呈现给你:

[L1] Controller | 任务分级:X 级

① 目标理解与拆分
   核心需求:
   拆解子目标:
   边界与约束:

② 技术路径与资源调度
   L2 链:
   L3 SubAgent:
   Skill 覆盖:(缺失:)
   风险项: 项

③ 缺口与需协助事项
   ❌ 缺失 Skill:[缺失清单] → 影响/建议
   ⚠️ 需你决策:[决策点]

④ 工作计划和节点
   

⑤ 任务编排
   执行顺序:
   风险与兜底:[风险清单]

第 5 步:等待确认

  • 你确认 → 进入执行阶段(第 6 步)
  • 你要求调整 → 重新生成或修改参数后重跑
  • 你否决 → 结束

第 6 步:调用 Workflow( controller-workflow ) 执行

收到确认后,调用 Workflow 工具执行完整任务编排:

Workflow({
  name: "controller-workflow",
  args: {
    userRequest: "",
    prePlan: {
      level: "",
      agentChain: "",
      l3Breakdown: ""
    }
  }
})

Workflow 执行过程中将依次显示:

  • [L1] Controller: 任务分级判断中...
  • [L2] Planner: 产品域工作中...(仅 S/A 级需要)
  • [L2] Architect: 研发域工作中...
  • [L3] Coder: 编码中... (节点 N)
  • [L3] Reviewer: 审查中... (节点 N)
  • [L3] Tester: 测试中... (节点 N)
  • [L2] Deployer: 运维域工作中...(仅 S 级需要)

重要规则

  1. 规划阶段不编码 — pre-planning-orchestrator 只输出方案,不写代码
  2. 必须等待你书面确认后才进入执行阶段
  3. 你确认后 controller-workflow 全权执行,中间不再打断
  4. B 级以下任务可简化:pre-planning 会跳过技能审计和可行性评估阶段
  5. C 级任务 pre-planning 直接返回,无需走 controller-workflow 执行

任务分级参考

| 级别 | 适用场景 | 规划深度 | 执行链 | |:----:|:---------|:---------|:-------| | S | 新项目、架构变更 | 完整 4 阶段 | planner→architect→deployer | | A | 多文件、有决策 | Agent分解+审计+评估 | planner→architect | | B | 单文件修改 | Agent 分解(跳过审计/评估) | architect | | C | 纯执行 | 直接返回 | 无 |

文件依赖

  • AGENT_REGISTRY.json — Agent 注册表,传给 pre-planning-orchestrator
  • CLAUDE.md — 技能清单,用于扫描 availableSkills
  • .claude/workflows/pre-planning-orchestrator.js — 规划阶段 Workflow 脚本
  • .claude/workflows/controller-workflow.js — 执行阶段 Workflow 脚本

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet, be the first.

Versions

  • v0.1.0 Imported from the upstream source.