How do I install the Pace skill?

Run npx skillmds@latest add wenjunduan/pace-6 in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Pace?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

wenjunduan (@wenjunduan) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Pace by wenjunduan — Agent Skill

Name: Pace
Author: wenjunduan

Pace

PACE 路由 + 8 stage 状态机. Athena 项目的中枢. 主 agent 进入任何 sprint 前必读. v9.7.0: 热路径精简 (铁律[三原语]). 本文件只保留路由判断; stage 详解 / 编排路由表 / hook 联动下沉 references/, 按需 Read.

wenjunduan 92b68de 19 files · 32.0 KB Updated Sep 17, 2026

File contents

PACE — Router & State Machine (v9.7.0, Codex)

6 路径

主 agent 收到用户输入后, 按改动量 + 紧急度判定路径:

路径	触发	stage 流程	强制 review?	强制 polish?	强制 worktree?
Hotfix	生产事故, 几分钟修	impl → ship	❌	❌	❌
Bugfix	已知 bug, 单文件	plan → impl → review → ship	✅ 单维	❌	❌
Quick	小改动, ≤3 文件	plan → impl → review → ship	✅ 单维	❌	❌
Feature	新功能, 单模块	plan → impl → review → ship	✅ 单维	❌	❌ (可选)
Refactor	改架构, ≥5 文件	plan → impl → review → polish → ship	✅ 三维度	✅	✅ 强制
System	跨模块, 系统级	plan → design → impl → review → polish → ship	✅ 三维度	✅	✅ 强制

8 Stage 状态机

                          (大需求或描述模糊时)
                                ↓
[brainstorm] ──→ [roadmap] ──→ plan ──→ [design] ──→ impl ──→ review ──→ [polish] ──→ ship
                                ↑           (System)              (3 维度)         (Refactor/System)
                                xhigh + critic 多轮

路由判断 (在 athena-dev 入口)

def route(user_input, ai_state):
    # 1. 显式信号
    if explicit_kws(["想法不清楚", "先 brainstorm", "讨论"]):
        return "brainstorm"
    if explicit_kws(["路线图", "分步推进", "拆分"]):
        return "roadmap"
    if explicit_kws(["bug", "修复"]):
        return "Bugfix"
    if explicit_kws(["重构", "refactor"]):
        return "Refactor"

    # 2. 隐式: 单词级模糊 → brainstorm (铁律[分诊])
    if len(user_input.split()) < 8 and not has_concrete_verb(user_input):
        return "brainstorm"

    # 3. ≥ 3 模块需求 → roadmap (铁律[分诊])
    if mentions_modules(user_input) >= 3:
        return "roadmap"

    # 4. 默认: 按改动量分 Feature/Quick/Hotfix
    return classify_by_scope(user_input)

写入路由 (铁律[零写入] 红黄绿区)

区	条件	执行者
绿	单文件 ≤30 行无跨模块影响, 或 Hotfix/Quick	主 thread 直接做
黄	单模块 Feature/Bugfix	spawn_agent, worktree 可选
红	Refactor/System 或并行 ≥2 写者	`git worktree add` + `spawn_agent --cwd` 强制

References (按需 Read, 不要预加载)

场景	Read
进入某 stage 前看详细工作流 / 数据目录	`references/stages.md`
选编排机制 (spawn_agent / multi-agent v2 / Goals)	`references/orchestration.md`
查 hook 联动 / compound 联动 / 项目级例外	`references/hooks.md`

最小循环提醒

plan/design: plan_mode_reasoning_effort = xhigh 已生效; critic 多轮 (max = _index.plan_critique_max_rounds)
impl: 按红黄绿区路由写入; PostToolUse(Bash) hook 自动写 evidence.yaml (过程证据)
review: 并行 reviewer + spec-compliance, 后跑 evaluator → VERDICT 写 _index.next_action
ship: delivery-gate 强制门禁 (cleanup-pass / architecture / re-review / spec-compliance)

---name: pace-6description: PACE 路由 + 8 stage 状态机. Athena 项目的中枢. 主 agent 进入任何 sprint 前必读. v9.7.0: 热路径精简 (铁律[三原语]). 本文件只保留路由判断; stage 详解 / 编排路由表 / hook 联动下沉 references/, 按需 Read.---# PACE — Router & State Machine (v9.7.0, Codex)## 6 路径主 agent 收到用户输入后, 按改动量 + 紧急度判定路径:| 路径 | 触发 | stage 流程 | 强制 review? | 强制 polish? | 强制 worktree? ||---|---|---|---|---|---|| **Hotfix** | 生产事故, 几分钟修 | impl → ship | ❌ | ❌ | ❌ || **Bugfix** | 已知 bug, 单文件 | plan → impl → review → ship | ✅ 单维 | ❌ | ❌ || **Quick** | 小改动, ≤3 文件 | plan → impl → review → ship | ✅ 单维 | ❌ | ❌ || **Feature** | 新功能, 单模块 | plan → impl → review → ship | ✅ 单维 | ❌ | ❌ (可选) || **Refactor** | 改架构, ≥5 文件 | plan → impl → review → polish → ship | ✅ 三维度 | ✅ | ✅ 强制 || **System** | 跨模块, 系统级 | plan → design → impl → review → polish → ship | ✅ 三维度 | ✅ | ✅ 强制 |## 8 Stage 状态机``` (大需求或描述模糊时) ↓[brainstorm] ──→ [roadmap] ──→ plan ──→ [design] ──→ impl ──→ review ──→ [polish] ──→ ship ↑ (System) (3 维度) (Refactor/System) xhigh + critic 多轮```## 路由判断 (在 athena-dev 入口)```pythondef route(user_input, ai_state): # 1. 显式信号 if explicit_kws(["想法不清楚", "先 brainstorm", "讨论"]): return "brainstorm" if explicit_kws(["路线图", "分步推进", "拆分"]): return "roadmap" if explicit_kws(["bug", "修复"]): return "Bugfix" if explicit_kws(["重构", "refactor"]): return "Refactor" # 2. 隐式: 单词级模糊 → brainstorm (铁律[分诊]) if len(user_input.split()) < 8 and not has_concrete_verb(user_input): return "brainstorm" # 3. ≥ 3 模块需求 → roadmap (铁律[分诊]) if mentions_modules(user_input) >= 3: return "roadmap" # 4. 默认: 按改动量分 Feature/Quick/Hotfix return classify_by_scope(user_input)```## 写入路由 (铁律[零写入] 红黄绿区)| 区 | 条件 | 执行者 ||---|---|---|| 绿 | 单文件 ≤30 行无跨模块影响, 或 Hotfix/Quick | 主 thread 直接做 || 黄 | 单模块 Feature/Bugfix | spawn_agent, worktree 可选 || 红 | Refactor/System 或并行 ≥2 写者 | `git worktree add` + `spawn_agent --cwd` 强制 |## References (按需 Read, 不要预加载)| 场景 | Read ||---|---|| 进入某 stage 前看详细工作流 / 数据目录 | `references/stages.md` || 选编排机制 (spawn_agent / multi-agent v2 / Goals) | `references/orchestration.md` || 查 hook 联动 / compound 联动 / 项目级例外 | `references/hooks.md` |## 最小循环提醒- plan/design: `plan_mode_reasoning_effort = xhigh` 已生效; critic 多轮 (max = `_index.plan_critique_max_rounds`)- impl: 按红黄绿区路由写入; PostToolUse(Bash) hook 自动写 evidence.yaml (过程证据)- review: 并行 reviewer + spec-compliance, 后跑 evaluator → VERDICT 写 `_index.next_action`- ship: delivery-gate 强制门禁 (cleanup-pass / architecture / re-review / spec-compliance)

Frequently asked questions