2 Commits
Author SHA1 Message Date
zhuyongxin 54741dd6c0 clean obsolete skill workspace files 2026-05-21 20:17:15 +08:00
zhuyongxin 6a5520db33 upgrade sm-flow v3.1 workflow 2026-05-21 19:55:23 +08:00
39 changed files with 955 additions and 1206 deletions
+36 -9
View File
@@ -1,4 +1,4 @@
---
---
name: sm-flow
description: OpenSpec-first、Devflow-assisted 的结构化工程开发元工作流。用户想把粗略想法、issue、PRD 或已有 research 推进为准确 OpenSpec change,并通过 OpenSpec apply 实现、验证、归档时使用;也适用于标准化开发流程、增强 OpenSpec 产物质量、沉淀工程决策和为未来代理保留上下文。
---
@@ -17,6 +17,8 @@ SM Flow 是一套 **OpenSpec-first / Devflow-assisted** 的软件交付元工作
- `openspec/changes/<change>/` 是执行真理源:当前变更的 proposal、design、specs 和 tasks。
- 代码是实现结果:只能在执行真理源足够明确后修改。
- 如果 `devflow/` 和 OpenSpec 冲突,不要直接写代码;先汇报冲突、让用户确认,并修正 OpenSpec。
- Phase 1 产出的 OpenSpec 默认为 **Draft OpenSpec**:它是澄清和审计对象,不是 Phase 3 的执行许可。
- 只有通过 Phase 2.9 提交检查后的 OpenSpec 才是 **Committed OpenSpec**;Phase 3 只能执行 Committed OpenSpec。
## 核心规则
@@ -25,16 +27,39 @@ SM Flow 是一套 **OpenSpec-first / Devflow-assisted** 的软件交付元工作
- devflow 默认采用分层产物:只强制保留人类理解和 OpenSpec 校准所必需的档案,避免重复 OpenSpec 的 proposal/design/tasks。
- 不得跳过 Phase 0.5:生成或修正 OpenSpec 前,必须读取相关 devflow 上下文。
- 不得跳过 Phase 2。即使需求看起来很清楚,也至少解决三个高价值澄清或验证问题。
- 不得跳过 Phase 2.9:进入 Phase 3 前,必须确认 Draft OpenSpec 已提交为 Committed OpenSpec。
- `evidence-driven` 不是自动通过:凡是通过证据确认的结论,也必须向用户汇报证据、结论和是否需要确认。
- `user-interview` 问题必须等待用户确认;涉及范围、偏好、验收口径、风险接受度时不能擅自决定。
- grill 过程必须显式 human-in-the-loop:每个 `user-interview` 问题一次只问一个,必须等用户回答并记录确认后才能继续;未确认的问题不能视为已解决。
- 单个 grill 决策确认不等于 Phase 3 apply 授权;Phase 2/2.5 回写 OpenSpec 后必须停在 Phase 2.9,报告 Committed OpenSpec 检查结果,并等待用户明确要求进入 apply / 开始实现 / 执行修改 / 继续 Phase 3。
- 接口影响按 L1-L4 分级处理;接口内部判断逻辑如果改变调用方可观察行为,也属于接口影响,不能因签名未变而降级。
- Phase 3 实现期冲突必须先分类再处理:实现偏差修代码;规格遗漏修 OpenSpec;设计冲突暂停并请用户确认;用户变更更新 proposal/specs/tasks 后再继续。
- 架构审计、澄清、ADR 或上下文发现如果影响实现,必须先回写 OpenSpec design/specs/tasks,再进入 Phase 3。
- 不得猜测式修 bug。实现失败或行为不明时,先进入 diagnose 闭环;如果根因是规格不准,先修 OpenSpec,再继续 apply。
- 默认采用 human-in-the-loop:Phase 1 后、Phase 2.5 后、Phase 3 前必须给用户一个简短检查点,除非用户明确要求“全自动执行”。
- 默认采用 human-in-the-loop:Phase 1 后、Phase 2.5 后、Phase 2.9 后、Phase 3 前必须给用户一个简短检查点;除非用户在启动时明确要求“全自动执行”,否则不得把单个决策确认推断为执行授权。
- Phase 4 结束前必须询问用户是否归档 OpenSpec change;不要默认执行归档。
- 子 skill 必须显式调用或显式降级:进入 Phase 1、1.5、2、2.5、3、4 时,先说明“本阶段调用/加载的 skill 是 X”;如果不能调用,必须说明“降级为 fallback”,并记录降级原因。
- 显式调用优先级:优先使用平台原生 Skill 调用;如果平台没有 Skill 工具,则读取对应 `SKILL.md` 并按其协议执行;只有文件不存在或协议不可执行时,才使用 `references/fallbacks.md`。
- 子 skill 绑定能力契约,不绑定单一平台路径。显式调用优先级:优先使用平台原生 Skill 调用;如果平台没有 Skill 工具,则读取对应能力的本地 `SKILL.md` 并按其协议执行;只有文件不存在或协议不可执行时,才使用 `references/fallbacks.md`。
- fallback 产物必须标注为降级执行,不能伪装成真实子 skill 调用。
## 接口影响分级
接口影响分级判断的是“记录在哪里、是否需要独立文档”,不是判断“是否需要关注”。凡涉及字段、DTO、service 方法、API、事件、回调、数据库契约、命令契约、跨模块调用语义或接口内部判断逻辑变化,都必须先判断接口影响。
| 级别 | 判断条件 | 产物要求 |
| --- | --- | --- |
| L1 内部实现 | 不改变任何调用方可观察的接口、字段、状态、错误码、数据范围、排序、过滤、权限结果、状态流转、副作用或文档承诺 | 不需要接口影响文档,只在 OpenSpec tasks 或 acceptance 记录验证 |
| L2 内部接口 | 改 DTO、service 方法、内部事件、内部 RPC 或内部判断逻辑,且所有消费者都在同一实现范围内 | 必须记录接口影响范围,可内联到 OpenSpec design/specs/tasks 或 devflow evidence/decisions |
| L3 协作接口 | 影响其他模块、其他服务、前端、外部系统、跨团队消费者、数据库契约、消息事件、回调或 SDK | 必须产出独立接口文档或等价独立章节 |
| L4 破坏性接口 | 删除字段、改字段语义、改状态机、改错误码、破坏兼容、旧调用方可能失败,或需要迁移、灰度、回滚 | 独立接口文档 + 迁移/回滚说明;必要时创建 ADR |
判断策略:
- 如果只是修复 bug,让接口回到原 OpenSpec 或原文档承诺,通常是 L1/L2。
- 如果判断逻辑改变了返回数据、错误码、状态、权限结果、排序/过滤、幂等性、时序或副作用,至少按 L3 检查。
- 如果旧调用方不改代码会失败、少数据、多数据、状态不同或错误码不同,按 L4 处理。
- 如果无法确定调用方边界或兼容性,默认提高一级并作为 `user-interview` 问题等待确认。
## 首次加载
执行前只读取当前任务需要的 reference 文件:
@@ -59,8 +84,8 @@ SM Flow 是一套 **OpenSpec-first / Devflow-assisted** 的软件交付元工作
- `devflow/reference/`
3. 如果根目录存在旧 `CONTEXT.md`,且 `devflow/glossary/CONTEXT.md` 不存在或为空,询问用户是迁移还是合并。
4. 检查 OpenSpec 和子 skill 是否可用:
- OpenSpec:`.claude/skills/openspec-propose/SKILL.md`、`.claude/skills/openspec-apply-change/SKILL.md`、`.claude/skills/openspec-archive-change/SKILL.md`。
- 辅助技能:`.agents/skills/to-prd/SKILL.md`、`.agents/skills/grill-with-docs/SKILL.md`、`.agents/skills/diagnose/SKILL.md`、`.agents/skills/tdd/SKILL.md`、`.agents/skills/zoom-out/SKILL.md`。
- OpenSpec 能力:`openspec-propose`、`openspec-apply-change`、`openspec-archive-change`。可由平台原生 skill、`.claude/skills/.../SKILL.md` 或 fallback 协议提供。
- 辅助能力:`to-prd`、`grill-with-docs`、`diagnose`、`tdd`、`zoom-out`。可由平台原生 skill、`.agents/skills/.../SKILL.md` 或 fallback 协议提供。
5. 如果 OpenSpec skill 不可用,不要直接绕过执行;先使用 fallback 生成/修正 OpenSpec 产物,并在 Phase 3 前向用户说明执行入口降级风险。
## 项目标识
@@ -101,13 +126,14 @@ devflow 是辅助 OpenSpec 和人类阅读的档案层,不复制 OpenSpec 的
## 阶段总览
1. Phase 0 — 入口澄清:接收初始 PRD、粗略想法或已有 research,澄清到可生成 OpenSpec。
2. Phase 0.5 — Devflow 上下文收集:读取 glossary、相关 PRD、ADR、acceptance、compound knowledge。
3. Phase 1 — OpenSpec propose:基于需求 + devflow 上下文生成或修正 OpenSpec proposal/design/specs/tasks。
2. Phase 0.5 — Devflow 上下文收集:先查 `devflow/index.md`,再读取 glossary、相关 PRD、ADR、acceptance、compound knowledge。
3. Phase 1 — OpenSpec propose:基于需求 + devflow 上下文生成或修正 Draft OpenSpec proposal/design/specs/tasks。
4. Phase 1.5 — PRD / OpenSpec 对齐:检查 OpenSpec 是否覆盖 PRD、遵守 ADR、使用正确术语、具备可验收 specs。
5. Phase 2 — Human-in-the-loop 澄清:声明 `evidence-driven` / `user-interview`,所有结论都汇报,确认后回写 OpenSpec。
6. Phase 2.5 — 架构审计:审计结果如果影响实现,必须回写 OpenSpec design/tasks。
7. Phase 3 — OpenSpec apply:默认依赖 OpenSpec 执行;devflow 只作为上下文参考。
8. Phase 4 — 回填 Devflow:从 OpenSpec、实现结果和验证结果提炼长期档案,并询问是否 archive OpenSpec change。
7. Phase 2.9 — Commit OpenSpec:检查 Draft OpenSpec 是否达到可执行状态,提交为 Committed OpenSpec。
8. Phase 3 — OpenSpec apply:默认依赖 Committed OpenSpec 执行;devflow 只作为上下文参考。
9. Phase 4 — 回填 Devflow:从 OpenSpec、实现结果和验证结果提炼长期档案,更新 `devflow/index.md`,并询问是否 archive OpenSpec change。
每个阶段的进入条件、动作、输出和退出标准见 `references/phase-contracts.md`。
@@ -117,6 +143,7 @@ devflow 是辅助 OpenSpec 和人类阅读的档案层,不复制 OpenSpec 的
- Phase 0.5 最小上下文检查:至少检查 glossary 和相关 ADR。
- Phase 2 最小澄清:至少一个术语问题、一个边界问题、一个验收问题;evidence-driven 结论也要汇报。
- Phase 2.9 提交检查:即使快速模式也必须确认没有未解决用户问题、接口影响已记录、OpenSpec tasks/specs 可执行。
- Phase 3 仍以 OpenSpec tasks/specs 为执行依据。
- Phase 4 轻量归档:记录验收结果、OpenSpec 产物链接和是否归档。
@@ -17,6 +17,10 @@ devflow/projects/YYYY-MM-DD-{slug}/
- `decisions.md`
- `acceptance.md`
同时维护仓库级索引:
- `devflow/index.md`
按需创建以下扩展文件:
- `prd.md`
@@ -49,6 +53,24 @@ devflow/projects/YYYY-MM-DD-{slug}/
| diagnose 记录 | 根因、修复、回归验证 | `acceptance.md` |
| 词汇表更新 | 术语和业务规则 | `devflow/glossary/CONTEXT.md` |
| 可复用经验 | 持久工程知识 | `devflow/compound/YYYY-MM-DD-{type}-{slug}.md` |
| 项目索引 | 日期、slug、领域、关键词、关联 OpenSpec、状态 | `devflow/index.md` |
## 索引维护规则
`devflow/index.md` 是 Phase 0.5 的默认入口,Phase 4 回填时必须维护。
最小字段:
| 日期 | slug | 领域 | 关键词 | 关联 OpenSpec | 状态 |
| --- | --- | --- | --- | --- | --- |
规则:
- 每个 `devflow/projects/YYYY-MM-DD-{slug}/` 默认对应一行索引。
- Phase 4 新建或更新项目档案时,必须新增或更新对应行。
- 如果项目仍在进行,状态写 `active`;已验收但未 archive 写 `accepted-unarchived`;已 archive 写 `archived`;暂停写 `paused`。
- 关键词只放能帮助 Phase 0.5 定位的术语,不复制 brief 内容。
- 如果无法准确判断领域或状态,写 `unknown`,并在 `acceptance.md` 记录待补。
## 验收记录规则
@@ -99,6 +121,7 @@ OpenSpec archive 是显式 human-in-the-loop 动作。archive 前必须确认 de
Phase 4 结束时告诉用户:
- 创建或更新了哪些档案文件。
- `devflow/index.md` 是否已更新。
- 运行了哪些验证,并按静态验证、脚本验证、浏览器/人工验证、未验证分类。
- 还剩哪些风险或后续事项。
- 明确询问:是否现在 archive OpenSpec change?
+14 -6
View File
@@ -34,12 +34,20 @@
1. 阅读 `proposal.md`、`design.md`、`specs/**/*.md` 和 `tasks.md`。
2. 确认 OpenSpec 与 devflow 上下文没有未解决冲突。
3. 修改前先检查现有代码。
4. 一次实现一个 OpenSpec task 的纵向切片。
5. 用最窄但有效的命令验证每个切片。
6. 只有验证通过或明确记录原因后,才更新 task 状态。
7. 如果失败原因不确定,停止并进入 diagnose。
8. 如果 diagnose 证明规格不准,先修正 OpenSpec,再继续执行。
3. 确认没有未解决的 user-interview 问题、未判级接口影响或未提交的 Draft OpenSpec。
4. 确认 Phase 2.9 后已经获得用户明确的 Phase 3 apply 授权;单个 grill 决策确认不能替代 apply 授权。
5. 修改前先检查现有代码。
6. 一次实现一个 OpenSpec task 的纵向切片。
7. 如果用户质疑、代码发现、测试失败或运行行为与 OpenSpec 冲突,先分类:
- 实现偏差:OpenSpec 正确,代码偏离;修代码。
- 规格遗漏:OpenSpec 未覆盖真实边界、接口影响或验收;暂停执行,修正 OpenSpec 并重新提交后继续。
- 设计冲突:OpenSpec 与架构、ADR、历史验收或模块边界冲突;暂停并等待用户确认。
- 用户变更:用户改变目标、范围、验收或风险接受度;更新 proposal/specs/tasks 后继续。
8. 冲突分类、证据、用户确认和 OpenSpec 回写状态必须记录到 `decisions.md` 或 `acceptance.md`。
9. 用最窄但有效的命令验证每个切片。
10. 只有验证通过或明确记录原因后,才更新 task 状态。
11. 如果失败原因不确定,停止并进入 diagnose。
12. 如果 diagnose 证明规格不准,先修正 OpenSpec,再继续执行。
## PRD fallback
@@ -28,6 +28,8 @@
**进入条件**:Phase 0 已经有足够信息定位领域、项目或变更方向。
**动作**:
- 优先读取 `devflow/index.md`,按日期、slug、领域、关键词和关联 OpenSpec 定位候选项目。
- 如果 `devflow/index.md` 不存在,先从 `devflow/projects/` 现有目录初始化轻量索引,再继续本次上下文收集。
- 读取 `devflow/glossary/CONTEXT.md`,提取相关术语和业务规则。
- 搜索 `devflow/projects/` 中相关 PRD、design、tasks、acceptance 和 ADR。
- 搜索 `devflow/compound/` 中可复用 learning、trick、decision、explore。
@@ -36,6 +38,7 @@
**退出条件**:
- 已形成“OpenSpec 输入上下文摘要”。
- 已记录 `devflow/index.md` 的使用状态:已命中 / 已初始化 / 无相关条目。
- 已列出相关 ADR 和不能违反的历史决策。
- 已列出需要写入或修正 OpenSpec 的上下文点。
@@ -46,7 +49,7 @@
**进入条件**:Phase 0 + Phase 0.5 已经足够生成或修正 OpenSpec change。
**显式子 skill**:`openspec-propose`。进入本阶段必须先声明调用方式:原生 Skill 调用 / 读取 `.claude/skills/openspec-propose/SKILL.md` / fallback 降级。
**显式子 skill**:`openspec-propose`。进入本阶段必须先声明调用方式:原生 Skill 调用 / 读取本地 `openspec-propose` 的 `SKILL.md` / fallback 降级。
**动作**:
- 优先调用 `openspec-propose`。
@@ -64,7 +67,7 @@
- 关键假设已显式记录在 OpenSpec 或 research 中。
**输出**:
- OpenSpec proposal、design、specs 和 task list。
- Draft OpenSpec proposal、design、specs 和 task list。
**Human checkpoint**:
- 向用户简要说明 OpenSpec scope、关键假设、主要风险、devflow 上下文如何影响 OpenSpec。
@@ -85,11 +88,16 @@
- OpenSpec 是否遵守相关 ADR。
- specs 是否能表达可观察行为。
- tasks 是否能驱动实现,而不是泛泛描述。
- 检查是否涉及接口影响:
- 是否改变字段、DTO、service 方法、API、事件、回调、数据库契约、命令契约或跨模块调用语义。
- 接口内部判断逻辑是否改变调用方可观察行为,例如返回数据、状态、错误码、权限结果、过滤/排序、幂等性、时序或副作用。
- 按 L1/L2/L3/L4 记录接口影响等级;不确定时标记为 Phase 2 的 `user-interview` 问题。
- 如果发现不一致,优先修正 OpenSpec,而不是只修改 devflow 文档。
**退出条件**:
- `brief.md` 已覆盖背景、目标、范围和非目标;复杂需求存在独立 `prd.md` 或用户明确不需要 PRD。
- OpenSpec 与 PRD/devflow 上下文没有已知冲突。
- 涉及接口变更时,已记录接口影响等级和产物要求;不确定项已进入 Phase 2。
- 所有已知冲突已修正或等待用户决策。
**输出**:
@@ -110,6 +118,10 @@
- `user-interview`:问题涉及产品偏好、范围边界、验收口径、风险接受度或价值取舍;必须问用户并等待确认。
- 至少覆盖三个维度:术语、边界、验收。
- 一次只问一个 `user-interview` 问题。
- 每个 `user-interview` 问题必须等待用户显式回答,并把回答、决策和 OpenSpec 回写状态记录到 `decisions.md`;不能由代理代替用户确认。
- 未获得用户确认的 `user-interview` 问题必须保持未解决状态,不能进入 Phase 2.9 或 Phase 3。
- 单个 `user-interview` 的确认只能解除该问题本身的阻塞,不能被解释为进入 Phase 3 或修改执行目标文件的授权。
- 对接口影响等级、消费者边界或兼容性存在不确定时,必须作为 `user-interview` 问题等待用户确认。
- 如果澄清结果影响实现,必须回写 OpenSpec proposal/design/specs/tasks。
- 术语一旦确认,更新 `devflow/glossary/CONTEXT.md`。
- 对难以逆转、依赖上下文、源自真实权衡的决策创建 ADR。
@@ -118,6 +130,8 @@
- 至少解决三个高价值澄清或验证问题,并记录每个问题属于 `evidence-driven` 还是 `user-interview`。
- 所有 evidence-driven 结论已向用户汇报。
- 所有 user-interview 决策已获得用户确认。
- 没有未解决或代理代确认的 user-interview 问题。
- 没有未判级或未确认的接口影响问题。
- 影响实现的结论已回写 OpenSpec。
**输出**:
@@ -148,14 +162,46 @@
**Human checkpoint**:
- 用不超过五句话向用户说明架构风险、OpenSpec 修正点和实现计划。
- 询问是否进入 Phase 3 OpenSpec apply;用户明确要求“全自动执行”时可跳过等待。
- 询问是否进入 Phase 2.9 Commit OpenSpec;用户明确要求“全自动执行”时可跳过等待。
## Phase 2.9 — Commit OpenSpec
**进入条件**:
- Phase 2 已解决术语、边界、验收三个维度的高价值问题。
- 所有 `user-interview` 问题都已获得用户显式确认。
- Phase 2.5 架构审计已经完成,或快速模式下已记录跳过原因。
- Draft OpenSpec 已回写所有会影响实现的澄清、接口影响和架构审计结论。
**动作**:
- 检查 proposal 是否说明为什么做、做什么、范围和非目标。
- 检查 design 是否记录上下文约束、关键技术决策、架构风险和接口影响。
- 检查 specs 是否表达外部可观察行为,并覆盖验收口径。
- 检查 tasks 是否是可执行的纵向切片,而不是泛泛描述。
- 检查接口影响是否已按 L1/L2/L3/L4 判级;L3/L4 是否有独立接口文档或等价独立章节。
- 检查没有未汇报的 evidence-driven 结论,没有未确认的 user-interview 问题,没有 devflow/OpenSpec 冲突。
- 如果检查失败,返回 Phase 1、Phase 2 或 Phase 2.5 修正 Draft OpenSpec。
**退出条件**:
- Draft OpenSpec 已达到可执行状态,并记录为 Committed OpenSpec。
- Phase 3 所需的 proposal、design、specs 和 tasks 均存在且一致。
- 所有 Phase 3 preflight 风险已消除或明确记录为已接受。
**输出**:
- Committed OpenSpec 状态说明。
- Phase 3 preflight 检查结果,默认写入 `decisions.md` 或 `acceptance.md`。
**Human checkpoint**:
- 用不超过五句话说明 Committed OpenSpec 的范围、接口影响、剩余风险和执行计划。
- 询问是否进入 Phase 3 OpenSpec apply;除非用户在启动时明确要求“全自动执行”,必须等待用户明确说出进入 apply、开始实现、执行修改、继续 Phase 3 或等价授权。
- 不得把 Phase 2 的单个 grill 决策确认当作本 checkpoint 的授权。
## Phase 3 — OpenSpec apply
**进入条件**:
- `openspec/changes/<slug>/` 中 proposal/design/specs/tasks 已达到可执行状态。
- Phase 2.5 后已获得用户继续实现的确认,除非用户要求“全自动执行”。
- `openspec/changes/<slug>/` 中 proposal/design/specs/tasks 已通过 Phase 2.9,成为 Committed OpenSpec。
- Phase 2.9 后已获得用户明确的 Phase 3 apply 授权,除非用户在启动时要求“全自动执行”。
- devflow 与 OpenSpec 没有未解决冲突。
- 没有未解决的 user-interview 问题、未判级接口影响、未汇报 evidence-driven 结论或未接受架构风险。
**显式子 skill**:`openspec-apply-change`;遇到 bug/不确定行为时显式调用 `diagnose`;需要测试驱动时显式调用 `tdd`。进入本阶段必须先声明调用方式,不能静默 fallback。
@@ -163,6 +209,12 @@
- 优先调用 `openspec-apply-change`。
- 执行依据是 OpenSpec specs/tasks;devflow 只能作为上下文参考。
- 按 OpenSpec tasks 的纵向切片实现。
- 当用户质疑、用户要求修改、代码检查、测试失败或运行行为与 OpenSpec 冲突时,先分类再继续:
- 实现偏差:Committed OpenSpec 正确,代码没有按规格做;修代码,不改 OpenSpec。
- 规格遗漏:OpenSpec 没覆盖真实边界、接口影响、验收口径或外部可见行为;暂停 apply,修正 OpenSpec 后重新提交。
- 设计冲突:OpenSpec 与架构、ADR、历史验收、数据所有权或模块边界冲突;暂停并向用户确认设计方向。
- 用户变更:用户改变目标、范围、优先级、验收或风险接受度;更新 proposal/specs/tasks 后再继续。
- 冲突分类、证据、用户确认和 OpenSpec 回写状态必须记录到 `decisions.md` 或 `acceptance.md`。
- 当用户要求、行为复杂或回归风险高时使用 TDD。
- 当测试失败、行为意外或原因不确定时使用 diagnose。
- 如果 diagnose 发现根因是 OpenSpec 不准确,先修正 OpenSpec,再继续 apply。
@@ -170,6 +222,7 @@
**退出条件**:
- OpenSpec tasks 已完成,或剩余 tasks 已明确记录。
- 所有实现期冲突已分类并处理;没有未确认的规格遗漏、设计冲突或用户变更。
- 已运行验证,或记录了未验证原因。
- 已列出已知限制。
@@ -188,10 +241,12 @@
- 从 OpenSpec、实现结果和验证结果提炼 brief、evidence、decisions 和 acceptance;只在复杂场景按需拆出 PRD/research/design/tasks/alignment。
- 写入或更新验收记录,并区分静态验证、脚本验证、浏览器/人工验证、未验证。
- 如果本次流程产生可复用经验,写入 compound knowledge。
- 更新 `devflow/index.md`,记录日期、slug、领域、关键词、关联 OpenSpec 和状态。
- 询问用户是否要 archive OpenSpec change;不要默认执行归档。
**退出条件**:
- `devflow/projects/YYYY-MM-DD-{slug}/` 能说明做了什么、为什么做、OpenSpec 如何指导执行、还剩什么、如何验证。
- `devflow/index.md` 已包含或更新本项目条目。
- 用户已被询问是否 archive OpenSpec change。
**输出**:
@@ -65,6 +65,68 @@
- 风险接受:{accepted by whom/when}
```
## 接口影响记录模板
```markdown
# {标题} 接口影响记录
## 分级
- 级别:L1 内部实现 / L2 内部接口 / L3 协作接口 / L4 破坏性接口
- 判级原因:{why this level}
- 是否需要独立接口文档:是 / 否
## 变更对象
- 接口/字段/DTO/事件/回调/数据库契约:
- 判断逻辑变化:
- 可观察行为变化:返回数据 / 状态 / 错误码 / 权限结果 / 过滤排序 / 幂等性 / 时序 / 副作用 / 无
## 影响范围
- 调用方/消费者:
- 是否跨模块/跨服务/跨团队:
- 旧调用方是否需要改动:
## 兼容与迁移
- 是否向后兼容:
- 迁移/灰度/回滚要求:
- 风险接受:
## 验收方式
- 如何证明新行为正确:
- 如何证明旧行为未破坏:
- 需要用户确认的问题:
```
## 实现期冲突记录模板
```markdown
# {标题} 实现期冲突记录
## 冲突摘要
- 触发来源:用户质疑 / 用户变更 / 代码发现 / 测试失败 / 运行行为
- 冲突对象:proposal / design / specs / tasks / ADR / 代码行为
- 分类:实现偏差 / 规格遗漏 / 设计冲突 / 用户变更
## 证据
- OpenSpec 依据:
- 代码或测试证据:
- 用户反馈:
## 处理
- 决策:
- 是否需要用户确认:是 / 否
- OpenSpec 回写:不需要 / 已回写 / 待回写
- 代码处理:
- 验证方式:
```
## PRD 模板
```markdown
@@ -1,136 +0,0 @@
# Skills v0.5.0 整合修复记录
**日期:** 2026-04-30
**来源:** grill-me 逐项审查,基于 v0.5.0 changelog 和 design doc 对比实际实现发现的不一致
---
## 一句话摘要
在前次 v0.5.0 slim architecture 改造的基础上,逐文件对比 design doc、proposal、plan、实际实现,修复了 `/essence` 遗漏项、`/follow` 旧引用残留、`/explore` 阶段冗余、references 死文件、docs 目录垃圾等问题。
---
## 已完成的改动
### 1. `/essence` — 补齐 v0.5.0 改造(之前遗漏)
| 位置 | 修改内容 |
|---|---|
| 版本号 | `0.3.0` → `0.5.0` |
| 透镜定义 | 删除 `(borrowed from /explore)`,声明为 `/essence` 自有 |
| 透镜相关 Phase 2/3/4/5 表格 | 从独立 3 列表格瘦身为内联 bullet point |
| HTML Card | `## Phase 6: Output (Optional HTML Card)` → `## Optional: HTML Card`;去 "production-critical" 条件,改为 `Only when the user explicitly requests it` |
| HTML Card 死链接 | 删除 `Same as /explore.` 两处,改为自包含内容 |
| Mode Selection 上下文感知 | 新增检测逻辑:来自 `/explore` → 默认 User-directed;独立启动 → 默认 Auto-detect |
| Auto-detect 信号 | `Most imported file` → `Cross-module contract`(后端项目更准确) |
### 2. `/follow` — 消除旧引用和设计偏差
| 位置 | 修改内容 |
|---|---|
| Mode Selection | 新增前置报告感知:来自 `/explore` + 代码仓库 → Runnable;来自 `/explore` + 非代码 → Reader;来自 `/essence` → Reader |
| Runnable Check | 删除重新扫描逻辑,改为从前置报告读取结论;运行时信号改为后端导向(`go.mod`, `pyproject.toml`, `Cargo.toml`, `Makefile` 等) |
| Reader Mode Flow | 从文件级 "Walk one core file" 提升为设计级 "Frame the learning goal around a core design" |
| 教学规则 | `do not use "go read the code" as the default instruction without context` → `never say "go read the code" as a standalone instruction`;引用代码必须先交代设计上下文 |
### 3. `/explore` — 合并冗余阶段
| 位置 | 修改内容 |
|---|---|
| Phase 1+2 | `Phase 1: Positioning` + `Phase 2: Structure` 合并为 `Phase 1: Positioning & Structure` |
| 阶段总数 | 5 Phase → 4 Phase |
| Project Type Detection | 信号列表同步(`package.json` 不再为首例);阶段跳过的编号更新 |
| Outcome 模板 | `5/5` → `4/4` |
### 4. References 清理
| 文件 | 操作 |
|---|---|
| `skills/explore/references/deep-fission.md` | **删除** — 描述已移除的 Deep Fission 功能,引用不存在的 `/fission` 技能 |
| `skills/essence/references/essence-signals.md` | `Most imported file` → `Cross-module contract`,与 SKILL.md 同步 |
### 5. docs 存档清理
| 文件 | 操作 | 原因 |
|---|---|---|
| `docs/superpowers/proposal.md` | 删除 | 原始 4-skill 提案,已过时 |
| `docs/superpowers/plan.md` | 删除 | 实施模板与实际实现不一致 |
| `docs/superpowers/plans/` | 删除 | 456 行 task-by-task 计划,含 `/map` 引用 |
| `docs/superpowers/README.md` | 删除 | `superpowers/README.md` 的副本,不属于历史存档 |
---
## docs 存档最终结构
```
docs/superpowers/
├── specs/
│ ├── issue.md # 原始需求
│ ├── 2026-04-21-skills-v0.5.0-slim-architecture-design.md
│ └── 2026-04-21-skills-v0.5.0-changelog-design.md
└── changelog/
├── 2026-04-21-skills-v0.5.0-slim-architecture.md # 第一次改造 changelog
├── 2026-04-21-skills-v0.5.0-validation-handoff.txt # 交接指令
└── 2026-04-30-skills-v0.5.0-consolidation.md # 本次整合(本文件)
```
---
## 当前 skill 状态
| 技能 | 版本 | Phase/Mode | 备注 |
|---|---|---|---|
| `/explore` | v0.5.0 | 4 Phase | 支持代码/非代码仓库;合并 Positioning+Structure |
| `/essence` | v0.5.0 | 5 Phase + Optional HTML | 透镜自有;上下文感知 Mode Selection |
| `/follow` | v0.5.0 | Runnable / Reader | 来源感知;不再重新扫描项目 |
---
## 实践验证结果
**日期:** 2026-04-30
**测试目标仓库:**
- 非代码仓库:explore-skill-family(自身)
- 代码仓库:SuperBizAgent-java(Spring Boot + AI Agent)
| # | 验证路径 | 目标 | 结果 |
|---|---|---|---|
| 1 | `/follow` 拒绝无前置报告 | 直接模拟 | ✅ |
| 2 | `/explore` 非代码仓库 | 自身 skill family | ✅ — 正确识别 skill-docs,跳过 Flow/Start |
| 3 | `/essence` 深挖 | "硬边界"设计 | ✅ — 结论-证据-解释,可迁移 |
| 4 | `/explore` → `/follow` | 自身 | ✅ — Reader 分流,引用报告内容 |
| 5 | `/essence` → `/follow` | 自身 | ✅ — design-focused,未重扫 |
| 6 | 三技能职责边界 | 综合矩阵检查 | ✅ — 核心问题/深度/前置依赖/产出均无重叠 |
| 7 | `/explore` 代码仓库 | SuperBizAgent-java | ✅ — 4 Phase 全流程,后端信号检测正常 |
### 验证发现的额外修复
验证过程中对 SKILL.md 的追加修改(已在代码中反映):
- `/essence` Phase 6 → Optional: HTML Card
- `/essence` Mode Selection 上下文感知
- `/essence` Most imported file → Cross-module contract
- `/follow` Mode Selection 来源感知
- `/follow` Runnable Check 不再重扫,引用前置报告
- `/follow` Reader Mode Flow 提升到设计层
- `/follow` Teaching Interaction Rules 收紧
- `/explore` Phase 1+2 合并,5→4 Phase
- `/explore` Project Type Detection 信号改为后端导向
- 删除 `skills/explore/references/deep-fission.md`
### handoff 10 条检查清单逐项结论
| # | 问题 | 结论 |
|---|---|---|
| 1 | `/explore` 对代码/非代码给出不同输出? | ✅ Phase 2/3 对非代码跳过 |
| 2 | `/explore` 保持在项目级理解? | ✅ Phase 4 概览深度 |
| 3 | `/essence` 稳定找最高价值设计? | ✅ "硬边界"提取到 pattern 层 |
| 4 | `/follow` 无前置报告明确拒绝? | ✅ 拒绝并引导 |
| 5 | `/follow` 基于 /explore 或 /essence 结果? | ✅ 两次测试均基于前置报告 |
| 6 | `/follow` 避免重新扫描? | ✅ 三次验证均未重扫 |
| 7 | `/follow` 真正引导式教学? | ✅ 先问后讲 |
| 8 | 三技能职责重叠/模糊? | ✅ 边界清晰 |
| 9 | README 与实际行为一致? | ✅ 三技能描述与 SKILL.md 一致 |
| 10 | references 与 v0.5.0 一致? | ✅ deep-fission 已删,signals 已同步 |
**总体验收结论:通过。** 文档、边界、行为三者一致。
+11
View File
@@ -0,0 +1,11 @@
# Devflow Index
`devflow/index.md` 是 Phase 0.5 的默认上下文入口。每个 `devflow/projects/YYYY-MM-DD-{slug}/` 默认维护一行,帮助代理先定位相关项目,再读取具体 brief、acceptance、ADR 或 compound knowledge。
| 日期 | slug | 领域 | 关键词 | 关联 OpenSpec | 状态 |
| --- | --- | --- | --- | --- | --- |
| 2026-05-18 | knowledge-index-panel | knowledge-index | knowledge-index, search, tags, static-html | `openspec/changes/knowledge-index-panel/` | active |
| 2026-05-19 | add-clear-filters | knowledge-index | clear-filters, search, tags, year-filter | `openspec/changes/archive/2026-05-19-add-clear-filters/` | archived |
| 2026-05-19 | add-year-filter | knowledge-index | year-filter, index-panel, frontmatter | `openspec/changes/add-year-filter/` | active |
| 2026-05-19 | dev-flow-skill-evaluation | sm-flow | dev-flow, sm-flow, skill-evaluation, fallback, phase-contracts | 未关联 | accepted-unarchived |
| 2026-05-21 | sm-flow-v3-1-upgrade | sm-flow | interface-impact, commit-gate, devflow-index, apply-conflict, skill-compatibility | `openspec/changes/archive/2026-05-21-sm-flow-v3-1-upgrade/` | archived |
@@ -0,0 +1,59 @@
# SM Flow v3.1 Upgrade Acceptance
## 结果
已接受。
## 验证
### 静态验证
- 命令/检查:`openspec status --change "sm-flow-v3-1-upgrade"`
- 结果:passed
- 备注:proposal、design、specs、tasks 均为 complete。
- 命令/检查:`rg -n "Draft OpenSpec|Committed OpenSpec|Phase 2.9|devflow/index.md|接口影响|实现期冲突|能力契约|平台原生 skill|user-interview" ...`
- 结果:passed
- 备注:关键术语已覆盖 `.agents/skills/sm-flow/`、OpenSpec change、workflow 文档和 devflow 项目档案。
### 脚本验证
- 命令:`openspec instructions apply --change "sm-flow-v3-1-upgrade" --json`
- 结果:passed
- 备注:所有 17 个 OpenSpec tasks 已完成。
### 浏览器/人工验证
- 步骤:不适用,本次是流程文档和 OpenSpec 规则改造。
- 结果:not run
- 备注:无 UI 行为。
## 已完成范围
- 增加 Draft OpenSpec / Committed OpenSpec / Phase 2.9 commit gate。
- 强化 Phase 2 grill 的显式人工确认约束。
- 增加“单个 grill 决策确认不等于 Phase 3 apply 授权”的硬约束。
- 增加接口影响 L1-L4 分级和接口影响记录模板。
- 增加 `devflow/index.md` 和 Phase 0.5 / Phase 4 索引维护规则。
- 增加 Phase 3 实现期冲突四分类和 fallback 执行规则。
- 将子 skill 兼容规则改为能力契约优先、路径其次。
- 更新 `skill-workbench/docs/sm-flow/workflow.md` 的 v3.1 说明。
## 已知限制
- `openspec/config.yaml` 由 OpenSpec CLI 创建并保持未跟踪状态。
- 工作区存在与本次无关的已删除文件,未处理。
## 流程偏差记录
- 偏差:本次执行中,Phase 2 grill 决策确认后曾直接进入执行文件修改,没有显式停在 Phase 2.9 checkpoint 等待 Phase 3 apply 授权。
- 分类:实现偏差。v3.1 规格方向正确,但执行过程没有严格遵守“grill 确认 ≠ apply 授权”的门禁。
- 修正:已回写 OpenSpec,新增 requirement 和 task,要求 Phase 2/2.5 回写后必须停在 Phase 2.9,并等待明确 Phase 3 apply 授权。
- 当前状态:用户已明确授权“执行apply”,该新增 requirement 已应用到 `.agents/skills/sm-flow/*` 和 workflow 文档。
## 交接
- 下一步:已归档,后续可从 devflow 档案和 OpenSpec archive 回溯。
- OpenSpec 归档确认:用户已确认归档。
- OpenSpec 归档位置:`openspec/changes/archive/2026-05-21-sm-flow-v3-1-upgrade/`
- Specs 同步:未执行;本仓库 `openspec/specs/` 当前无主 spec 文件,本次保留 delta specs 于 archive 中。
@@ -0,0 +1,30 @@
# SM Flow v3.1 Upgrade Brief
## 背景
- 用户目标:把 `sm-flow` v3 使用中暴露的接口文档、流程顺序、上下文定位、子 skill 兼容和实现期冲突问题固化为 v3.1 协议。
- 当前问题:v3 已经确立 OpenSpec-first,但缺少接口影响分级、Draft/Committed OpenSpec gate、devflow 索引和 Phase 3 冲突沟通规则。
- 关联 OpenSpec:`openspec/changes/sm-flow-v3-1-upgrade/`
- devflow 分档:standard
## 范围
- 本次要做:更新 `sm-flow` skill 本体、phase contracts、fallbacks、templates、workflow 说明,并创建/维护 devflow 索引。
- 本次不做:修改业务代码、重写 OpenSpec CLI、把 devflow 变成执行真理源、强制所有接口变更都独立产出接口文档。
- 影响区域:
- `.agents/skills/sm-flow/SKILL.md`
- `.agents/skills/sm-flow/references/*.md`
- `skill-workbench/docs/sm-flow/workflow.md`
- `devflow/index.md`
## OpenSpec 对齐
- proposal 覆盖状态:已覆盖
- specs 覆盖状态:已覆盖
- tasks 覆盖状态:已覆盖
## OpenSpec 输入上下文摘要
- v3 历史决策:`devflow/` 是上下文真理源,OpenSpec 是执行真理源,代码是实现结果。
- 用户问题来源:`skill-workbench/docs/sm-flow/使用问题.md`,共 6 个问题,集中在接口产物、流程顺序、上下文检索、子 skill 兼容和实现期冲突。
- 历史评估来源:`devflow/projects/2026-05-19-dev-flow-skill-evaluation/dev-flow-skill-evaluation.md`,已确认子 skill fallback、quick 模式、Phase 退出条件和 OpenSpec-first 是 v3 的关键设计。
@@ -0,0 +1,39 @@
# SM Flow v3.1 Upgrade Decisions
## User-interview
| 问题 | 用户回答 | 决策 | OpenSpec 回写 |
| --- | --- | --- | --- |
| 是否启动 v3.1 改造? | “开始这次的v3.1改造” | 创建 `sm-flow-v3-1-upgrade` OpenSpec change 并进入改造。 | 已回写 |
| grill 过程中是否必须人工确认? | 用户要求“如果 skill 里并没有强调 grill 的过程中一定要人工确认的话,加上这个约束”。 | Phase 2 的 user-interview 问题必须一问一答等待用户显式确认;未确认不能进入 Phase 2.9 / Phase 3。 | 已回写 |
| 接口变更是否需要记录影响范围? | 用户要求“只要涉及接口里的变更,不管是新增字段还是内部变更,都需要添加到接口影响范围中”。 | 所有接口变更必须有影响记录。 | 已回写 |
| 接口影响采用什么分级策略? | 用户确认“可以,先按照分级来实现”。 | 采用 L1-L4 分级:L1 内部实现、L2 内部接口、L3 协作接口、L4 破坏性接口;接口内部判断逻辑按可观察行为评估。 | 已回写 |
| devflow 上下文索引是否强制维护? | 用户确认“可以”。 | 强制维护轻量 `devflow/index.md`;Phase 0.5 先查 index,Phase 4 回填时必须更新 index。 | 已回写 |
| 实现阶段遇到与原设计冲突的质疑或修改时如何处理? | 用户确认“可以”。 | Phase 3 采用四类实现期冲突分类:实现偏差、规格遗漏、设计冲突、用户变更;先分类和确认,再继续代码或 OpenSpec 修正。 | 已回写 |
| 子 skill 兼容规则是否采用能力优先? | 用户确认“可以”。 | 子 skill 绑定能力契约,不绑定单一平台路径;调用优先级为平台原生 skill、本地 `SKILL.md`、sm-flow fallback。 | 已回写 |
| 单个 grill 决策确认是否等于 apply 授权? | 用户确认“可以”。 | 单个 `user-interview` 的“可以”只确认该决策;Phase 2/2.5 回写 OpenSpec 后必须停在 Phase 2.9,等待明确 Phase 3 apply 授权。 | 已回写 |
## Evidence-driven 澄清
| 维度 | 模式 | 问题 | 证据 / 用户反馈 | 结论 | 是否回写 OpenSpec |
| --- | --- | --- | --- | --- | --- |
| 术语 | evidence-driven | “接口影响范围文档”和“接口文档”是否同一产物? | 用户分别列为两个问题;流程需要降低文档重复。 | 区分“接口影响记录”和“独立接口文档”。 | 是 |
| 边界 | evidence-driven | v3.1 是否要改变 OpenSpec-first 主轴? | v3 skill 本体和历史评估都把 OpenSpec 定义为执行真理源。 | 不改变主轴,只补 gate。 | 是 |
| 验收 | evidence-driven | 如何证明 v3.1 改造完成? | OpenSpec specs 已覆盖接口影响、commit gate、上下文索引和 apply 冲突处理。 | 验收以文档规则可检索、OpenSpec status 完成、tasks 勾选为准。 | 是 |
## 关键取舍
- 决策:采用 Draft / Committed OpenSpec,而不是把 grill 移到 propose 前。
- 原因:Draft 给澄清提供结构化对象,Committed 防止草稿直接执行。
- 影响:新增 Phase 2.9,但减少 Phase 3 规格漂移。
- 风险接受:本次 v3.1 接受。
- 决策:接口变更采用 L1-L4 分级。
- 原因:所有接口影响都需要追踪,但只有高影响变更需要独立接口文档。
- 影响:templates 和 phase contracts 需要增加接口影响记录。
- 风险接受:本次 v3.1 接受。
- 决策:为 devflow 增加 `devflow/index.md`。
- 原因:项目档案增长后,需要索引而不是每次全量翻阅。
- 影响:Phase 0.5 和 Phase 4 都需要维护索引。
- 风险接受:本次 v3.1 接受。
@@ -0,0 +1,28 @@
# SM Flow v3.1 Upgrade Evidence
## 证据
| 来源 | 证据 | 结论 | 是否已汇报 |
| --- | --- | --- | --- |
| `skill-workbench/docs/sm-flow/使用问题.md` | 用户列出 6 个实践问题,包含接口文档、propose/grill 顺序、devflow 检索、子 skill 兼容、Phase 3 设计冲突。 | v3.1 应围绕这些使用摩擦改协议,而不是重写整个流程。 | 是 |
| `.agents/skills/sm-flow/SKILL.md` | 当前核心规则已要求 Phase 0.5、Phase 2、Phase 2.5、Phase 3、Phase 4,但没有 Draft/Committed OpenSpec 和接口影响分级。 | 新规则应补 gate,不应推翻 v3 的 OpenSpec-first 主从关系。 | 是 |
| `.agents/skills/sm-flow/references/phase-contracts.md` | Phase 1 生成 OpenSpec 后进入 Phase 1.5/2/2.5,Phase 3 以 OpenSpec 执行;缺少 Phase 2.9 提交检查。 | “propose 后 grill 返工”应通过草稿/提交分离解释和治理。 | 是 |
| `.agents/skills/sm-flow/references/fallbacks.md` | 执行 fallback 已要求失败原因不确定时 diagnose,规格不准先修 OpenSpec。 | 可扩展为 Phase 3 实现期冲突分类,不需要新建完全独立流程。 | 是 |
| `devflow/projects/2026-05-19-dev-flow-skill-evaluation/dev-flow-skill-evaluation.md` | 历史评估已指出 Claude 工具耦合、子 skill 调用方式不稳、Phase 退出条件不足。 | v3.1 的兼容规则应绑定能力契约,而不是绑定具体平台路径。 | 是 |
## Evidence-driven 结论
- 结论:v3.1 应保留 v3 主轴,只增加 gate 和分级规则。
- 证据:当前 skill 文件和 phase contracts 已经清楚表达 OpenSpec-first。
- 风险:如果重写阶段顺序,容易引入新的执行歧义。
- 用户确认:已通过“开始这次的 v3.1 改造”确认推进。
- 结论:接口文档应分级,不应一刀切。
- 证据:用户同时提出“接口影响范围文档”和“接口文档”,说明需要区分影响记录与独立文档。
- 风险:分级阈值不清会导致代理低估外部契约风险。
- 用户确认:需要在实现时以不确定则确认为规则。
- 结论:Phase 3 应增加实现期沟通规则,而不是让代码建议直接覆盖 OpenSpec。
- 证据:用户明确要求“不要全部接收代码建议,而是向我确认是设计冲突,还是没有考虑到”。
- 风险:沟通 gate 会暂停执行,但能保护规格真理源。
- 用户确认:已确认这是 v3.1 重点。
@@ -0,0 +1,2 @@
schema: spec-driven
created: 2026-05-21
@@ -0,0 +1,54 @@
## Context
SM Flow v3 已经把 `devflow/` 定位为上下文真理源,把 `openspec/changes/<change>/` 定位为执行真理源。当前摩擦来自执行细节:一些需要人类判断的规则没有形成 gate,导致代理可能过早把 Draft OpenSpec 当成最终规格,或者在 Phase 3 中用代码侧发现直接覆盖原设计。
本次改造的对象是 skill 协议本身,主要文件位于 `.agents/skills/sm-flow/`,历史说明位于 `skill-workbench/docs/sm-flow/workflow.md`。用户提供的问题记录位于 `skill-workbench/docs/sm-flow/使用问题.md`。
## Goals / Non-Goals
**Goals:**
- 让 v3.1 明确“文档不是越多越好”:devflow 记录上下文、证据、决策和验收,OpenSpec 记录执行依据。
- 对接口变更建立分级规则,避免“所有接口变更都独立成文档”和“接口影响没人记录”两个极端。
- 引入 Draft / Committed OpenSpec,允许 Phase 1 先形成讨论对象,但禁止未提交的草稿直接进入 Phase 3。
- 为 devflow 增加索引入口和固定检索顺序,解决项目档案增长后的定位问题。
- 在 Phase 3 增加实现期沟通和冲突分类规则,避免把用户质疑、测试失败或代码建议直接当作新规格。
**Non-Goals:**
- 不重写 OpenSpec CLI 或 `.claude/skills/openspec-*`。
- 不改业务代码或知识索引功能。
- 不把 devflow 重新提升为执行真理源。
- 不强制每次接口变更都创建独立接口文档。
## Decisions
1. **Draft / Committed OpenSpec 分离**
- 决策:Phase 1 产物称为 Draft OpenSpec;Phase 2 和 Phase 2.5 后进入新增的 Phase 2.9 Commit OpenSpec,只有通过提交检查的 OpenSpec 才能进入 Phase 3。
- 原因:完全推迟 OpenSpec 会缺少讨论对象;完全信任 Phase 1 初稿又会让 grill 后返工显得像异常。草稿/提交分离把返工变成正常流程。
- 替代方案:把 grill 放到 propose 前。拒绝原因是缺少结构化规格草稿时,澄清容易停留在对话层,难以精确回写 proposal/specs/tasks。
2. **接口影响分级,而不是固定独立文档**
- 决策:所有接口变更都必须有接口影响记录;只有 L3/L4 级别才必须独立产出接口文档。
- 原因:接口变更需要可追踪,但低风险内部接口不应制造额外文档负担。
- 替代方案:凡接口变更都新增接口文档。拒绝原因是会让 micro/standard 变更过重,增加文档重复和漂移。
3. **devflow 通过索引和检索顺序控制增长**
- 决策:新增或维护 `devflow/index.md`,并规定 Phase 0.5 的检索顺序为 `devflow/index.md`、`devflow/glossary/CONTEXT.md`、相关项目 brief/acceptance/ADR、compound knowledge。
- 原因:devflow 的长期价值来自可回溯;没有索引时,文档越多越像一片温柔但很黏的沼泽。
- 替代方案:每次用全文搜索全量扫。拒绝原因是成本随项目数增长,且容易抓到无关历史。
4. **Phase 3 冲突先分类,再执行**
- 决策:实现阶段遇到用户质疑、代码建议、测试失败或新事实与 OpenSpec 冲突时,必须分类为实现偏差、规格遗漏、设计冲突或用户变更。
- 原因:Phase 3 的职责是执行已提交规格,不是把所有新输入立即吸收成代码改动。
- 替代方案:让代理自行判断并继续。拒绝原因是会破坏 OpenSpec 的执行真理源地位。
5. **子 skill 绑定能力契约**
- 决策:文档应表达 `openspec-propose`、`openspec-apply-change` 等能力名和调用优先级;`.claude/skills/...` 是一个实现路径,不是唯一前提。
- 原因:同一流程应能迁移到 Codex、Claude 或其他代理环境。
- 替代方案:固定 Claude 路径。拒绝原因是兼容性弱,且与当前 `.agents/skills/` 运行方式不匹配。
## Risks / Trade-offs
- Draft / Committed OpenSpec 增加一个 Phase 2.9 gate → 用固定检查清单控制成本,避免变成新一轮大文档。
- 接口影响分级可能被代理误判 → 把分级阈值写成可观察条件,并要求不确定时向用户确认。
- `devflow/index.md` 需要维护 → Phase 4 回填时把索引更新列为默认动作,减少遗忘。
- Phase 3 冲突分类会暂停执行 → 这是刻意设计;暂停一次比悄悄把规格改歪更便宜。
@@ -0,0 +1,34 @@
## Why
SM Flow v3 已经确立了 OpenSpec-first / Devflow-assisted 的主从关系,但实际使用中仍存在四类摩擦:接口变更文档粒度不清、propose 后再 grill 导致返工、devflow 增长后上下文定位变慢、实现阶段发现设计冲突时容易被代码建议牵着走。
这次 v3.1 改造要把这些摩擦固化为可执行规则,让代理在生成规格、澄清需求、执行 apply 和回填 devflow 时有明确 gate,而不是靠临场判断。
## What Changes
- 增加接口影响分级规则:所有接口变更都必须记录影响,只有达到跨团队、外部契约或破坏性变更阈值时才独立产出接口文档。
- 增加 OpenSpec 草稿/提交分离:Phase 1 生成 Draft OpenSpec,Phase 2/2.5 澄清和审计后通过 Phase 2.9 提交为 Phase 3 的执行依据。
- 增加 devflow 上下文定位规则:通过 `devflow/index.md`、项目 `brief.md` 和固定检索顺序减少全量翻阅。
- 增加实现期冲突处理规则:Phase 3 中用户质疑、测试失败、代码发现和 OpenSpec 冲突时,必须先分类再继续执行。
- 增加 Phase 3 前 OpenSpec 可执行性检查,确保 proposal/design/specs/tasks、接口影响、未解决用户问题和 devflow 冲突都达标。
- 调整子 skill 兼容规则:把流程绑定到能力契约,而不是绑定到 Claude 路径;Claude skill 文件只是一个可用实现。
- 更新 sm-flow skill 本体、phase contracts、fallbacks、templates 和工作流说明文档。
## Capabilities
### New Capabilities
- `sm-flow-interface-impact`: 定义接口影响分级、接口影响记录和独立接口文档的触发条件。
- `sm-flow-commit-gate`: 定义 Draft OpenSpec、Committed OpenSpec、Phase 2.9 提交检查和 Phase 3 前可执行性 gate。
- `sm-flow-context-indexing`: 定义 devflow 上下文索引和固定检索顺序。
- `sm-flow-apply-conflict-handling`: 定义 Phase 3 实现期冲突分类、用户确认和 OpenSpec 回写规则。
### Modified Capabilities
<!-- 没有已归档的 openspec/specs 需要修改;当前仓库尚未建立全局 specs 目录。 -->
## Impact
- 修改 `.agents/skills/sm-flow/SKILL.md`。
- 修改 `.agents/skills/sm-flow/references/phase-contracts.md`、`fallbacks.md`、`templates.md`。
- 可能新增 `devflow/index.md` 作为长期上下文索引入口。
- 更新 `skill-workbench/docs/sm-flow/workflow.md`,记录 v3.1 的最终设计。
- 不修改业务代码,不改变 `knowledge/` 索引功能。
@@ -0,0 +1,39 @@
## ADDED Requirements
### Requirement: Phase 3 classifies implementation-time conflicts
SM Flow SHALL classify implementation-time conflicts before changing code or OpenSpec.
#### Scenario: User challenges implementation during apply
- **WHEN** the user questions or changes implementation behavior during Phase 3 and the request conflicts with Committed OpenSpec
- **THEN** the flow classifies the issue as implementation deviation, spec omission, design conflict, or user scope change before continuing
#### Scenario: Code suggests a different approach
- **WHEN** code inspection, tests, or runtime behavior suggests a different approach than Committed OpenSpec
- **THEN** the flow reports the evidence and classification instead of silently accepting the code-side suggestion
#### Scenario: Conflict is implementation deviation
- **WHEN** Committed OpenSpec remains correct and implementation diverges from proposal, design, specs, or tasks
- **THEN** the flow classifies the issue as implementation deviation and fixes the code without changing executable OpenSpec except task status or acceptance notes
#### Scenario: Conflict is spec omission
- **WHEN** Committed OpenSpec lacks a real boundary, behavior, validation rule, interface impact, or acceptance case discovered during implementation
- **THEN** the flow classifies the issue as spec omission and returns to OpenSpec repair before continuing apply
#### Scenario: Conflict is design conflict
- **WHEN** Committed OpenSpec conflicts with architecture, ADR, historical acceptance, data ownership, lifecycle, or module boundaries
- **THEN** the flow classifies the issue as design conflict, pauses implementation, reports the conflict, and asks the user to confirm the design direction
#### Scenario: Conflict is user scope change
- **WHEN** the user changes the goal, scope, priority, acceptance expectation, or risk tolerance during implementation
- **THEN** the flow classifies the issue as user scope change and updates proposal, specs, and tasks before continuing
### Requirement: Spec-affecting conflicts return to OpenSpec repair
SM Flow SHALL repair OpenSpec before continuing when implementation-time conflicts affect the executable specification.
#### Scenario: Conflict is a spec omission or design conflict
- **WHEN** a Phase 3 conflict changes scope, externally observable behavior, interface compatibility, architecture decisions, or task slicing
- **THEN** the flow pauses apply, obtains required user confirmation, updates proposal/design/specs/tasks, reruns the commit gate, and only then resumes Phase 3
#### Scenario: Conflict is implementation deviation
- **WHEN** the committed specification is still correct and the code diverges from it
- **THEN** the flow fixes the implementation without changing OpenSpec except for task status or acceptance notes
@@ -0,0 +1,49 @@
## ADDED Requirements
### Requirement: OpenSpec drafts are not executable
SM Flow SHALL distinguish Draft OpenSpec from Committed OpenSpec.
#### Scenario: Phase 1 proposal created
- **WHEN** Phase 1 creates or updates proposal, design, specs, and tasks
- **THEN** the flow treats those artifacts as Draft OpenSpec until Phase 2, Phase 2.5, and Phase 2.9 checks are complete
#### Scenario: Draft has unresolved questions
- **WHEN** Draft OpenSpec contains unresolved user-interview questions, unreported evidence-driven conclusions, architecture conflicts, or unrecorded interface impact
- **THEN** the flow SHALL NOT enter Phase 3
### Requirement: Phase 2.9 commits executable OpenSpec
SM Flow SHALL include a Phase 2.9 Commit OpenSpec gate before Phase 3.
#### Scenario: Commit gate passes
- **WHEN** proposal explains scope and non-goals, design captures implementation constraints, specs describe observable behavior, tasks are executable vertical slices, interface impact is recorded, and devflow conflicts are resolved
- **THEN** the flow marks the OpenSpec change as Committed OpenSpec and may ask the user to proceed to Phase 3
#### Scenario: Commit gate fails
- **WHEN** any required executable OpenSpec condition is missing
- **THEN** the flow returns to Phase 1, Phase 2, or Phase 2.5 to repair the missing artifact before implementation
### Requirement: Grill questions require explicit human confirmation
SM Flow SHALL treat Phase 2 grill as a human-in-the-loop process that requires explicit user confirmation for user-interview questions.
#### Scenario: User-interview grill question is asked
- **WHEN** Phase 2 raises a user-interview question about terminology, scope, acceptance, risk, priority, or product preference
- **THEN** the flow asks exactly one question, waits for the user's answer, records the answer, and only then continues to the next user-interview question
#### Scenario: Grill confirmation is missing
- **WHEN** a user-interview question has no explicit user answer
- **THEN** the flow SHALL NOT mark the question resolved, SHALL NOT commit OpenSpec, and SHALL NOT enter Phase 3
### Requirement: Grill confirmation does not authorize apply
SM Flow SHALL distinguish confirmation of an individual grill decision from authorization to enter Phase 3.
#### Scenario: User confirms a grill decision
- **WHEN** the user answers a Phase 2 user-interview question with confirmation such as "可以", "确认", or equivalent
- **THEN** the flow records that decision and updates OpenSpec/devflow, but SHALL NOT treat the answer as authorization to modify execution target files
#### Scenario: OpenSpec has been updated after grill
- **WHEN** Phase 2 or Phase 2.5 findings have been written back to proposal, design, specs, or tasks
- **THEN** the flow stops at Phase 2.9, reports the Committed OpenSpec check result, and waits for explicit user authorization to enter Phase 3
#### Scenario: Apply authorization is missing
- **WHEN** the user has not explicitly said to enter apply, start implementation, execute the changes, continue Phase 3, or equivalent
- **THEN** the flow may update OpenSpec and devflow decision records, but SHALL NOT modify execution target files
@@ -0,0 +1,23 @@
## ADDED Requirements
### Requirement: Devflow context lookup uses an index-first strategy
SM Flow SHALL use a stable index-first devflow lookup order to find relevant context.
#### Scenario: Phase 0.5 starts
- **WHEN** Phase 0.5 collects devflow context
- **THEN** the flow checks `devflow/index.md` first, then `devflow/glossary/CONTEXT.md`, then related project brief/acceptance/ADR files, then `devflow/compound/`
#### Scenario: Index is missing during lookup
- **WHEN** `devflow/index.md` does not exist
- **THEN** the flow initializes a lightweight index from known project directories, falls back to targeted search for the current lookup, and records that the index was bootstrapped
### Requirement: Phase 4 maintains devflow index
SM Flow SHALL update devflow index metadata during Phase 4 when a project archive is created or updated.
#### Scenario: Project is backfilled
- **WHEN** Phase 4 creates or updates `devflow/projects/YYYY-MM-DD-{slug}/`
- **THEN** the flow updates `devflow/index.md` with date, slug, domain, keywords, related OpenSpec change, and status
#### Scenario: Phase 4 completes without index update
- **WHEN** Phase 4 has created or updated project backfill files but has not updated `devflow/index.md`
- **THEN** the flow is incomplete and must update the index before reporting completion
@@ -0,0 +1,46 @@
## ADDED Requirements
### Requirement: Interface changes have impact records
SM Flow SHALL require every interface-related change to record interface impact before Phase 3.
#### Scenario: Internal interface field changes
- **WHEN** a change adds, removes, renames, or changes the semantics of a field, DTO, service method, event, API, callback, database contract, or command contract
- **THEN** the flow records the affected interface, affected consumers, compatibility expectation, and validation method in OpenSpec design/specs/tasks or devflow evidence/decisions
#### Scenario: Interface uncertainty
- **WHEN** the agent cannot determine whether a change affects an interface contract
- **THEN** the flow treats it as an interface-impact question and asks for confirmation before Phase 3
#### Scenario: Internal decision logic changes observable behavior
- **WHEN** a change modifies internal decision logic inside an interface and the result can change returned data, status, error code, permission result, validation result, ordering, filtering, idempotency, timing, or side effects
- **THEN** the flow treats the change as interface impact even if the interface shape and field names are unchanged
### Requirement: Interface documentation is level-gated
SM Flow SHALL create an independent interface document only when the interface impact level requires it.
#### Scenario: Low-risk internal interface change
- **WHEN** an interface change is limited to internal implementation or an internal module boundary and has no external consumers
- **THEN** the flow records the interface impact inline without requiring a standalone interface document
#### Scenario: External or breaking interface change
- **WHEN** an interface change affects external APIs, SDKs, callbacks, events, database contracts, cross-team consumers, or breaks backward compatibility
- **THEN** the flow requires a standalone interface document or equivalent explicit section covering consumers, compatibility, migration, rollback, and validation
### Requirement: Interface impact levels use risk-based classification
SM Flow SHALL classify interface impact by consumer boundary, contract semantics, and compatibility risk.
#### Scenario: L1 internal implementation
- **WHEN** a change does not alter any consumer-visible interface shape, field, status, error, data range, ordering, permission result, state transition, side effect, or documented behavior
- **THEN** the flow classifies it as L1 and records validation in tasks or acceptance without requiring interface impact documentation
#### Scenario: L2 internal interface
- **WHEN** a change affects internal DTOs, service methods, internal events, internal RPC, or internal decision logic and all consumers are inside the same implementation scope
- **THEN** the flow classifies it as L2 and records interface impact inline
#### Scenario: L3 collaboration interface
- **WHEN** a change affects other modules, services, frontend callers, external systems, cross-team consumers, database contracts, events, callbacks, or SDK users
- **THEN** the flow classifies it as L3 and requires a standalone interface document or equivalent explicit section
#### Scenario: L4 breaking interface
- **WHEN** old consumers can fail, receive less data, receive more data, observe different statuses or errors, require migration, require rollback, or lose backward compatibility
- **THEN** the flow classifies it as L4 and requires standalone interface documentation plus migration and rollback notes
@@ -0,0 +1,31 @@
## 1. OpenSpec Commit Gate
- [x] 1.1 Update `.agents/skills/sm-flow/SKILL.md` to describe Draft OpenSpec, Committed OpenSpec, Phase 2.9, and the Phase 3 executable gate.
- [x] 1.2 Update `.agents/skills/sm-flow/references/phase-contracts.md` with Phase 2.9 entry/action/output/exit criteria.
- [x] 1.3 Add Phase 3 preflight checks for unresolved user-interview questions, interface impact, devflow conflicts, and executable task/spec quality.
- [x] 1.4 Require Phase 2 grill user-interview questions to wait for explicit user confirmation before they can be marked resolved.
- [x] 1.5 Require explicit Phase 3 apply authorization after Phase 2.9; individual grill confirmations must not authorize execution target file changes.
## 2. Interface Impact Rules
- [x] 2.1 Add interface impact level definitions and inline-vs-standalone documentation rules to the sm-flow protocol.
- [x] 2.2 Add an interface impact template to `.agents/skills/sm-flow/references/templates.md`.
- [x] 2.3 Update Phase 1.5 / Phase 2 checks so interface changes are identified before Phase 3.
## 3. Devflow Context Indexing
- [x] 3.1 Add devflow index lookup order to Phase 0.5.
- [x] 3.2 Create or update `devflow/index.md` with existing project entries.
- [x] 3.3 Update Phase 4 rules so future project backfills maintain the index.
## 4. Apply Conflict Handling
- [x] 4.1 Add Phase 3 implementation-time conflict classification rules.
- [x] 4.2 Update `.agents/skills/sm-flow/references/fallbacks.md` so execution fallback pauses and repairs OpenSpec for spec-affecting conflicts.
- [x] 4.3 Ensure conflict classifications are recorded in devflow decisions or acceptance notes.
## 5. Compatibility and Documentation
- [x] 5.1 Reword sub-skill compatibility rules to bind to capability contracts first and implementation paths second.
- [x] 5.2 Update `skill-workbench/docs/sm-flow/workflow.md` with the final v3.1 design.
- [x] 5.3 Run OpenSpec status checks and text searches to verify the new terms are consistently documented.
+20
View File
@@ -0,0 +1,20 @@
schema: spec-driven
# Project context (optional)
# This is shown to AI when creating artifacts.
# Add your tech stack, conventions, style guides, domain knowledge, etc.
# Example:
# context: |
# Tech stack: TypeScript, React, Node.js
# We use conventional commits
# Domain: e-commerce platform
# Per-artifact rules (optional)
# Add custom rules for specific artifacts.
# Example:
# rules:
# proposal:
# - Keep proposals under 500 words
# - Always include a "Non-goals" section
# tasks:
# - Break tasks into chunks of max 2 hours
+92
View File
@@ -378,6 +378,98 @@ Phase 4 回填 devflow,并确认是否 archive
`sm-flow` 不是 OpenSpec 的替代品,而是 OpenSpec 的上下文增强层和归档闭环层。
## v3.1:可执行 gate 与使用摩擦修正
v3.1 来自一次真实使用复盘:接口变更文档粒度、propose 后再 grill 的返工感、devflow 增长后的检索成本、子 skill 路径兼容,以及 Phase 3 中用户质疑和代码发现与原设计冲突时的处理方式,都需要变成可执行规则。
v3.1 不改变 v3 的主轴:
> `devflow/` 提供上下文,OpenSpec 提供执行依据,代码只是实现结果。
### Draft / Committed OpenSpec
Phase 1 产出的 OpenSpec 统一视为 **Draft OpenSpec**。它是后续 PRD 对齐、grill 和架构审计的讨论对象,不是实现许可。
Phase 2 和 Phase 2.5 的结论必须回写 OpenSpec。随后新增 Phase 2.9:
```text
Phase 2.9 Commit OpenSpec
```
Phase 2.9 检查:
- proposal 是否说明范围、非目标和原因。
- design 是否记录关键约束、架构风险和接口影响。
- specs 是否覆盖可观察行为和验收口径。
- tasks 是否是可执行的纵向切片。
- 是否还有未确认的 user-interview、未汇报的 evidence-driven 结论、未判级接口影响或 devflow/OpenSpec 冲突。
只有通过 Phase 2.9 的 OpenSpec 才是 **Committed OpenSpec**,Phase 3 只能执行 Committed OpenSpec。
### Grill 必须人工确认
Phase 2 的 `grill-with-docs` 不是代理自问自答。`user-interview` 问题必须一次只问一个,等待用户显式回答,并记录到 `decisions.md`。未获得用户确认的问题不能视为已解决,也不能进入 Phase 2.9 或 Phase 3。
`evidence-driven` 问题可以由代理先查证,但也必须汇报证据、结论和是否需要用户确认。
单个 grill 决策确认不等于 Phase 3 apply 授权。用户在 Phase 2 回答“可以”只表示该问题已确认;代理必须先回写 OpenSpec/devflow,停在 Phase 2.9,报告 Committed OpenSpec 检查结果,并等待用户明确要求进入 apply、开始实现、执行修改或继续 Phase 3。
### 接口影响分级
v3.1 区分“接口影响记录”和“独立接口文档”:
- 所有接口变更都必须记录影响范围。
- 只有 L3/L4 才强制独立接口文档或等价独立章节。
分级规则:
| 级别 | 判断条件 | 产物要求 |
| --- | --- | --- |
| L1 内部实现 | 不改变调用方可观察行为 | tasks / acceptance 记录验证 |
| L2 内部接口 | 内部 DTO、service 方法、内部事件、内部判断逻辑变化,消费者仍在同一实现范围内 | 内联接口影响记录 |
| L3 协作接口 | 影响前端、其他模块/服务、跨团队消费者、数据库契约、事件、回调或 SDK | 独立接口文档或独立章节 |
| L4 破坏性接口 | 旧调用方可能失败、数据/状态/错误码不同,或需要迁移/灰度/回滚 | 独立接口文档 + 迁移/回滚说明,必要时 ADR |
接口内部判断逻辑也按可观察行为评估。即使签名和字段不变,只要返回数据、错误码、状态、权限结果、排序/过滤、幂等性、时序或副作用变化,也属于接口影响。
### Devflow Index
`devflow/index.md` 成为 Phase 0.5 的默认入口。上下文查找顺序为:
```text
devflow/index.md
devflow/glossary/CONTEXT.md
相关项目 brief / acceptance / ADR
devflow/compound/
```
Phase 4 回填项目档案时必须更新 `devflow/index.md`。最小字段为日期、slug、领域、关键词、关联 OpenSpec 和状态。
### 实现期冲突处理
Phase 3 中,如果用户质疑、用户修改、代码发现、测试失败或运行行为与 Committed OpenSpec 冲突,必须先分类:
| 分类 | 含义 | 处理 |
| --- | --- | --- |
| 实现偏差 | OpenSpec 正确,代码没有按规格做 | 修代码,不改 OpenSpec |
| 规格遗漏 | OpenSpec 没覆盖真实边界、接口影响、验收或外部行为 | 暂停 apply,修正 OpenSpec,重新 commit |
| 设计冲突 | OpenSpec 与架构、ADR、历史验收、数据所有权或模块边界冲突 | 暂停并请用户确认设计方向 |
| 用户变更 | 用户改变目标、范围、验收或风险接受度 | 更新 proposal/specs/tasks 后继续 |
冲突分类、证据、用户确认和 OpenSpec 回写状态必须进入 `decisions.md` 或 `acceptance.md`。
### 子 skill 兼容
v3.1 将子 skill 从“路径绑定”改为“能力绑定”。例如 `openspec-propose`、`openspec-apply-change`、`to-prd`、`grill-with-docs` 都是能力契约。
调用优先级:
1. 平台原生 skill。
2. 本地 `SKILL.md`,例如 `.claude/skills/...` 或 `.agents/skills/...`。
3. `sm-flow` 的 fallback 协议。
这样 `sm-flow` 可以在 Claude、Codex 或其他代理环境中迁移,而不会被某个固定目录锁死。
@@ -0,0 +1,8 @@
1. 如果涉及接口,需要额外产出一份影响接口范围文档
1. ❯ 只要涉及到接口里的变更,不管是新增字段还是内部变更,都需要添加到接口影响范围中
2. 涉及接口变更,需要产出一份接口文档
3. 顺序问题,目前的顺序是propose之后在,生成了task、design之后,再用grill去澄清需求,这样当澄清完需求后,又得回去更新task、design、proposal等文档,现在的流程,是propose之后,再通过to-prd,生成devflow的产物,然后再grill来更新,哪种设计更合理?
4. 如果devflow数量多了,如何快速定位需要的上下文
5. 子skill现在是固定了claude,要考虑兼容情况
6. 当在开发阶段,我发起质疑或者修改,和原来设计冲突时,不要全部接收代码建议,而是向我确认是设计冲突,还是没有考虑到,要提出反馈
1. 可能需要加多一个步骤,apply实施过程中的沟通?
@@ -1,4 +1,4 @@
---
---
name: sm-flow
description: OpenSpec-first、Devflow-assisted 的结构化工程开发元工作流。用户想把粗略想法、issue、PRD 或已有 research 推进为准确 OpenSpec change,并通过 OpenSpec apply 实现、验证、归档时使用;也适用于标准化开发流程、增强 OpenSpec 产物质量、沉淀工程决策和为未来代理保留上下文。
---
@@ -17,6 +17,8 @@ SM Flow 是一套 **OpenSpec-first / Devflow-assisted** 的软件交付元工作
- `openspec/changes/<change>/` 是执行真理源:当前变更的 proposal、design、specs 和 tasks。
- 代码是实现结果:只能在执行真理源足够明确后修改。
- 如果 `devflow/` 和 OpenSpec 冲突,不要直接写代码;先汇报冲突、让用户确认,并修正 OpenSpec。
- Phase 1 产出的 OpenSpec 默认为 **Draft OpenSpec**:它是澄清和审计对象,不是 Phase 3 的执行许可。
- 只有通过 Phase 2.9 提交检查后的 OpenSpec 才是 **Committed OpenSpec**;Phase 3 只能执行 Committed OpenSpec。
## 核心规则
@@ -25,16 +27,39 @@ SM Flow 是一套 **OpenSpec-first / Devflow-assisted** 的软件交付元工作
- devflow 默认采用分层产物:只强制保留人类理解和 OpenSpec 校准所必需的档案,避免重复 OpenSpec 的 proposal/design/tasks。
- 不得跳过 Phase 0.5:生成或修正 OpenSpec 前,必须读取相关 devflow 上下文。
- 不得跳过 Phase 2。即使需求看起来很清楚,也至少解决三个高价值澄清或验证问题。
- 不得跳过 Phase 2.9:进入 Phase 3 前,必须确认 Draft OpenSpec 已提交为 Committed OpenSpec。
- `evidence-driven` 不是自动通过:凡是通过证据确认的结论,也必须向用户汇报证据、结论和是否需要确认。
- `user-interview` 问题必须等待用户确认;涉及范围、偏好、验收口径、风险接受度时不能擅自决定。
- grill 过程必须显式 human-in-the-loop:每个 `user-interview` 问题一次只问一个,必须等用户回答并记录确认后才能继续;未确认的问题不能视为已解决。
- 单个 grill 决策确认不等于 Phase 3 apply 授权;Phase 2/2.5 回写 OpenSpec 后必须停在 Phase 2.9,报告 Committed OpenSpec 检查结果,并等待用户明确要求进入 apply / 开始实现 / 执行修改 / 继续 Phase 3。
- 接口影响按 L1-L4 分级处理;接口内部判断逻辑如果改变调用方可观察行为,也属于接口影响,不能因签名未变而降级。
- Phase 3 实现期冲突必须先分类再处理:实现偏差修代码;规格遗漏修 OpenSpec;设计冲突暂停并请用户确认;用户变更更新 proposal/specs/tasks 后再继续。
- 架构审计、澄清、ADR 或上下文发现如果影响实现,必须先回写 OpenSpec design/specs/tasks,再进入 Phase 3。
- 不得猜测式修 bug。实现失败或行为不明时,先进入 diagnose 闭环;如果根因是规格不准,先修 OpenSpec,再继续 apply。
- 默认采用 human-in-the-loop:Phase 1 后、Phase 2.5 后、Phase 3 前必须给用户一个简短检查点,除非用户明确要求“全自动执行”。
- 默认采用 human-in-the-loop:Phase 1 后、Phase 2.5 后、Phase 2.9 后、Phase 3 前必须给用户一个简短检查点;除非用户在启动时明确要求“全自动执行”,否则不得把单个决策确认推断为执行授权。
- Phase 4 结束前必须询问用户是否归档 OpenSpec change;不要默认执行归档。
- 子 skill 必须显式调用或显式降级:进入 Phase 1、1.5、2、2.5、3、4 时,先说明“本阶段调用/加载的 skill 是 X”;如果不能调用,必须说明“降级为 fallback”,并记录降级原因。
- 显式调用优先级:优先使用平台原生 Skill 调用;如果平台没有 Skill 工具,则读取对应 `SKILL.md` 并按其协议执行;只有文件不存在或协议不可执行时,才使用 `references/fallbacks.md`。
- 子 skill 绑定能力契约,不绑定单一平台路径。显式调用优先级:优先使用平台原生 Skill 调用;如果平台没有 Skill 工具,则读取对应能力的本地 `SKILL.md` 并按其协议执行;只有文件不存在或协议不可执行时,才使用 `references/fallbacks.md`。
- fallback 产物必须标注为降级执行,不能伪装成真实子 skill 调用。
## 接口影响分级
接口影响分级判断的是“记录在哪里、是否需要独立文档”,不是判断“是否需要关注”。凡涉及字段、DTO、service 方法、API、事件、回调、数据库契约、命令契约、跨模块调用语义或接口内部判断逻辑变化,都必须先判断接口影响。
| 级别 | 判断条件 | 产物要求 |
| --- | --- | --- |
| L1 内部实现 | 不改变任何调用方可观察的接口、字段、状态、错误码、数据范围、排序、过滤、权限结果、状态流转、副作用或文档承诺 | 不需要接口影响文档,只在 OpenSpec tasks 或 acceptance 记录验证 |
| L2 内部接口 | 改 DTO、service 方法、内部事件、内部 RPC 或内部判断逻辑,且所有消费者都在同一实现范围内 | 必须记录接口影响范围,可内联到 OpenSpec design/specs/tasks 或 devflow evidence/decisions |
| L3 协作接口 | 影响其他模块、其他服务、前端、外部系统、跨团队消费者、数据库契约、消息事件、回调或 SDK | 必须产出独立接口文档或等价独立章节 |
| L4 破坏性接口 | 删除字段、改字段语义、改状态机、改错误码、破坏兼容、旧调用方可能失败,或需要迁移、灰度、回滚 | 独立接口文档 + 迁移/回滚说明;必要时创建 ADR |
判断策略:
- 如果只是修复 bug,让接口回到原 OpenSpec 或原文档承诺,通常是 L1/L2。
- 如果判断逻辑改变了返回数据、错误码、状态、权限结果、排序/过滤、幂等性、时序或副作用,至少按 L3 检查。
- 如果旧调用方不改代码会失败、少数据、多数据、状态不同或错误码不同,按 L4 处理。
- 如果无法确定调用方边界或兼容性,默认提高一级并作为 `user-interview` 问题等待确认。
## 首次加载
执行前只读取当前任务需要的 reference 文件:
@@ -59,8 +84,8 @@ SM Flow 是一套 **OpenSpec-first / Devflow-assisted** 的软件交付元工作
- `devflow/reference/`
3. 如果根目录存在旧 `CONTEXT.md`,且 `devflow/glossary/CONTEXT.md` 不存在或为空,询问用户是迁移还是合并。
4. 检查 OpenSpec 和子 skill 是否可用:
- OpenSpec:`.claude/skills/openspec-propose/SKILL.md`、`.claude/skills/openspec-apply-change/SKILL.md`、`.claude/skills/openspec-archive-change/SKILL.md`。
- 辅助技能:`.agents/skills/to-prd/SKILL.md`、`.agents/skills/grill-with-docs/SKILL.md`、`.agents/skills/diagnose/SKILL.md`、`.agents/skills/tdd/SKILL.md`、`.agents/skills/zoom-out/SKILL.md`。
- OpenSpec 能力:`openspec-propose`、`openspec-apply-change`、`openspec-archive-change`。可由平台原生 skill、`.claude/skills/.../SKILL.md` 或 fallback 协议提供。
- 辅助能力:`to-prd`、`grill-with-docs`、`diagnose`、`tdd`、`zoom-out`。可由平台原生 skill、`.agents/skills/.../SKILL.md` 或 fallback 协议提供。
5. 如果 OpenSpec skill 不可用,不要直接绕过执行;先使用 fallback 生成/修正 OpenSpec 产物,并在 Phase 3 前向用户说明执行入口降级风险。
## 项目标识
@@ -101,13 +126,14 @@ devflow 是辅助 OpenSpec 和人类阅读的档案层,不复制 OpenSpec 的
## 阶段总览
1. Phase 0 — 入口澄清:接收初始 PRD、粗略想法或已有 research,澄清到可生成 OpenSpec。
2. Phase 0.5 — Devflow 上下文收集:读取 glossary、相关 PRD、ADR、acceptance、compound knowledge。
3. Phase 1 — OpenSpec propose:基于需求 + devflow 上下文生成或修正 OpenSpec proposal/design/specs/tasks。
2. Phase 0.5 — Devflow 上下文收集:先查 `devflow/index.md`,再读取 glossary、相关 PRD、ADR、acceptance、compound knowledge。
3. Phase 1 — OpenSpec propose:基于需求 + devflow 上下文生成或修正 Draft OpenSpec proposal/design/specs/tasks。
4. Phase 1.5 — PRD / OpenSpec 对齐:检查 OpenSpec 是否覆盖 PRD、遵守 ADR、使用正确术语、具备可验收 specs。
5. Phase 2 — Human-in-the-loop 澄清:声明 `evidence-driven` / `user-interview`,所有结论都汇报,确认后回写 OpenSpec。
6. Phase 2.5 — 架构审计:审计结果如果影响实现,必须回写 OpenSpec design/tasks。
7. Phase 3 — OpenSpec apply:默认依赖 OpenSpec 执行;devflow 只作为上下文参考。
8. Phase 4 — 回填 Devflow:从 OpenSpec、实现结果和验证结果提炼长期档案,并询问是否 archive OpenSpec change。
7. Phase 2.9 — Commit OpenSpec:检查 Draft OpenSpec 是否达到可执行状态,提交为 Committed OpenSpec。
8. Phase 3 — OpenSpec apply:默认依赖 Committed OpenSpec 执行;devflow 只作为上下文参考。
9. Phase 4 — 回填 Devflow:从 OpenSpec、实现结果和验证结果提炼长期档案,更新 `devflow/index.md`,并询问是否 archive OpenSpec change。
每个阶段的进入条件、动作、输出和退出标准见 `references/phase-contracts.md`。
@@ -117,6 +143,7 @@ devflow 是辅助 OpenSpec 和人类阅读的档案层,不复制 OpenSpec 的
- Phase 0.5 最小上下文检查:至少检查 glossary 和相关 ADR。
- Phase 2 最小澄清:至少一个术语问题、一个边界问题、一个验收问题;evidence-driven 结论也要汇报。
- Phase 2.9 提交检查:即使快速模式也必须确认没有未解决用户问题、接口影响已记录、OpenSpec tasks/specs 可执行。
- Phase 3 仍以 OpenSpec tasks/specs 为执行依据。
- Phase 4 轻量归档:记录验收结果、OpenSpec 产物链接和是否归档。
@@ -17,6 +17,10 @@ devflow/projects/YYYY-MM-DD-{slug}/
- `decisions.md`
- `acceptance.md`
同时维护仓库级索引:
- `devflow/index.md`
按需创建以下扩展文件:
- `prd.md`
@@ -49,6 +53,24 @@ devflow/projects/YYYY-MM-DD-{slug}/
| diagnose 记录 | 根因、修复、回归验证 | `acceptance.md` |
| 词汇表更新 | 术语和业务规则 | `devflow/glossary/CONTEXT.md` |
| 可复用经验 | 持久工程知识 | `devflow/compound/YYYY-MM-DD-{type}-{slug}.md` |
| 项目索引 | 日期、slug、领域、关键词、关联 OpenSpec、状态 | `devflow/index.md` |
## 索引维护规则
`devflow/index.md` 是 Phase 0.5 的默认入口,Phase 4 回填时必须维护。
最小字段:
| 日期 | slug | 领域 | 关键词 | 关联 OpenSpec | 状态 |
| --- | --- | --- | --- | --- | --- |
规则:
- 每个 `devflow/projects/YYYY-MM-DD-{slug}/` 默认对应一行索引。
- Phase 4 新建或更新项目档案时,必须新增或更新对应行。
- 如果项目仍在进行,状态写 `active`;已验收但未 archive 写 `accepted-unarchived`;已 archive 写 `archived`;暂停写 `paused`。
- 关键词只放能帮助 Phase 0.5 定位的术语,不复制 brief 内容。
- 如果无法准确判断领域或状态,写 `unknown`,并在 `acceptance.md` 记录待补。
## 验收记录规则
@@ -99,6 +121,7 @@ OpenSpec archive 是显式 human-in-the-loop 动作。archive 前必须确认 de
Phase 4 结束时告诉用户:
- 创建或更新了哪些档案文件。
- `devflow/index.md` 是否已更新。
- 运行了哪些验证,并按静态验证、脚本验证、浏览器/人工验证、未验证分类。
- 还剩哪些风险或后续事项。
- 明确询问:是否现在 archive OpenSpec change?
@@ -34,12 +34,20 @@
1. 阅读 `proposal.md`、`design.md`、`specs/**/*.md` 和 `tasks.md`。
2. 确认 OpenSpec 与 devflow 上下文没有未解决冲突。
3. 修改前先检查现有代码。
4. 一次实现一个 OpenSpec task 的纵向切片。
5. 用最窄但有效的命令验证每个切片。
6. 只有验证通过或明确记录原因后,才更新 task 状态。
7. 如果失败原因不确定,停止并进入 diagnose。
8. 如果 diagnose 证明规格不准,先修正 OpenSpec,再继续执行。
3. 确认没有未解决的 user-interview 问题、未判级接口影响或未提交的 Draft OpenSpec。
4. 确认 Phase 2.9 后已经获得用户明确的 Phase 3 apply 授权;单个 grill 决策确认不能替代 apply 授权。
5. 修改前先检查现有代码。
6. 一次实现一个 OpenSpec task 的纵向切片。
7. 如果用户质疑、代码发现、测试失败或运行行为与 OpenSpec 冲突,先分类:
- 实现偏差:OpenSpec 正确,代码偏离;修代码。
- 规格遗漏:OpenSpec 未覆盖真实边界、接口影响或验收;暂停执行,修正 OpenSpec 并重新提交后继续。
- 设计冲突:OpenSpec 与架构、ADR、历史验收或模块边界冲突;暂停并等待用户确认。
- 用户变更:用户改变目标、范围、验收或风险接受度;更新 proposal/specs/tasks 后继续。
8. 冲突分类、证据、用户确认和 OpenSpec 回写状态必须记录到 `decisions.md` 或 `acceptance.md`。
9. 用最窄但有效的命令验证每个切片。
10. 只有验证通过或明确记录原因后,才更新 task 状态。
11. 如果失败原因不确定,停止并进入 diagnose。
12. 如果 diagnose 证明规格不准,先修正 OpenSpec,再继续执行。
## PRD fallback
@@ -28,6 +28,8 @@
**进入条件**:Phase 0 已经有足够信息定位领域、项目或变更方向。
**动作**:
- 优先读取 `devflow/index.md`,按日期、slug、领域、关键词和关联 OpenSpec 定位候选项目。
- 如果 `devflow/index.md` 不存在,先从 `devflow/projects/` 现有目录初始化轻量索引,再继续本次上下文收集。
- 读取 `devflow/glossary/CONTEXT.md`,提取相关术语和业务规则。
- 搜索 `devflow/projects/` 中相关 PRD、design、tasks、acceptance 和 ADR。
- 搜索 `devflow/compound/` 中可复用 learning、trick、decision、explore。
@@ -36,6 +38,7 @@
**退出条件**:
- 已形成“OpenSpec 输入上下文摘要”。
- 已记录 `devflow/index.md` 的使用状态:已命中 / 已初始化 / 无相关条目。
- 已列出相关 ADR 和不能违反的历史决策。
- 已列出需要写入或修正 OpenSpec 的上下文点。
@@ -46,7 +49,7 @@
**进入条件**:Phase 0 + Phase 0.5 已经足够生成或修正 OpenSpec change。
**显式子 skill**:`openspec-propose`。进入本阶段必须先声明调用方式:原生 Skill 调用 / 读取 `.claude/skills/openspec-propose/SKILL.md` / fallback 降级。
**显式子 skill**:`openspec-propose`。进入本阶段必须先声明调用方式:原生 Skill 调用 / 读取本地 `openspec-propose` 的 `SKILL.md` / fallback 降级。
**动作**:
- 优先调用 `openspec-propose`。
@@ -64,7 +67,7 @@
- 关键假设已显式记录在 OpenSpec 或 research 中。
**输出**:
- OpenSpec proposal、design、specs 和 task list。
- Draft OpenSpec proposal、design、specs 和 task list。
**Human checkpoint**:
- 向用户简要说明 OpenSpec scope、关键假设、主要风险、devflow 上下文如何影响 OpenSpec。
@@ -85,11 +88,16 @@
- OpenSpec 是否遵守相关 ADR。
- specs 是否能表达可观察行为。
- tasks 是否能驱动实现,而不是泛泛描述。
- 检查是否涉及接口影响:
- 是否改变字段、DTO、service 方法、API、事件、回调、数据库契约、命令契约或跨模块调用语义。
- 接口内部判断逻辑是否改变调用方可观察行为,例如返回数据、状态、错误码、权限结果、过滤/排序、幂等性、时序或副作用。
- 按 L1/L2/L3/L4 记录接口影响等级;不确定时标记为 Phase 2 的 `user-interview` 问题。
- 如果发现不一致,优先修正 OpenSpec,而不是只修改 devflow 文档。
**退出条件**:
- `brief.md` 已覆盖背景、目标、范围和非目标;复杂需求存在独立 `prd.md` 或用户明确不需要 PRD。
- OpenSpec 与 PRD/devflow 上下文没有已知冲突。
- 涉及接口变更时,已记录接口影响等级和产物要求;不确定项已进入 Phase 2。
- 所有已知冲突已修正或等待用户决策。
**输出**:
@@ -110,6 +118,10 @@
- `user-interview`:问题涉及产品偏好、范围边界、验收口径、风险接受度或价值取舍;必须问用户并等待确认。
- 至少覆盖三个维度:术语、边界、验收。
- 一次只问一个 `user-interview` 问题。
- 每个 `user-interview` 问题必须等待用户显式回答,并把回答、决策和 OpenSpec 回写状态记录到 `decisions.md`;不能由代理代替用户确认。
- 未获得用户确认的 `user-interview` 问题必须保持未解决状态,不能进入 Phase 2.9 或 Phase 3。
- 单个 `user-interview` 的确认只能解除该问题本身的阻塞,不能被解释为进入 Phase 3 或修改执行目标文件的授权。
- 对接口影响等级、消费者边界或兼容性存在不确定时,必须作为 `user-interview` 问题等待用户确认。
- 如果澄清结果影响实现,必须回写 OpenSpec proposal/design/specs/tasks。
- 术语一旦确认,更新 `devflow/glossary/CONTEXT.md`。
- 对难以逆转、依赖上下文、源自真实权衡的决策创建 ADR。
@@ -118,6 +130,8 @@
- 至少解决三个高价值澄清或验证问题,并记录每个问题属于 `evidence-driven` 还是 `user-interview`。
- 所有 evidence-driven 结论已向用户汇报。
- 所有 user-interview 决策已获得用户确认。
- 没有未解决或代理代确认的 user-interview 问题。
- 没有未判级或未确认的接口影响问题。
- 影响实现的结论已回写 OpenSpec。
**输出**:
@@ -148,14 +162,46 @@
**Human checkpoint**:
- 用不超过五句话向用户说明架构风险、OpenSpec 修正点和实现计划。
- 询问是否进入 Phase 3 OpenSpec apply;用户明确要求“全自动执行”时可跳过等待。
- 询问是否进入 Phase 2.9 Commit OpenSpec;用户明确要求“全自动执行”时可跳过等待。
## Phase 2.9 — Commit OpenSpec
**进入条件**:
- Phase 2 已解决术语、边界、验收三个维度的高价值问题。
- 所有 `user-interview` 问题都已获得用户显式确认。
- Phase 2.5 架构审计已经完成,或快速模式下已记录跳过原因。
- Draft OpenSpec 已回写所有会影响实现的澄清、接口影响和架构审计结论。
**动作**:
- 检查 proposal 是否说明为什么做、做什么、范围和非目标。
- 检查 design 是否记录上下文约束、关键技术决策、架构风险和接口影响。
- 检查 specs 是否表达外部可观察行为,并覆盖验收口径。
- 检查 tasks 是否是可执行的纵向切片,而不是泛泛描述。
- 检查接口影响是否已按 L1/L2/L3/L4 判级;L3/L4 是否有独立接口文档或等价独立章节。
- 检查没有未汇报的 evidence-driven 结论,没有未确认的 user-interview 问题,没有 devflow/OpenSpec 冲突。
- 如果检查失败,返回 Phase 1、Phase 2 或 Phase 2.5 修正 Draft OpenSpec。
**退出条件**:
- Draft OpenSpec 已达到可执行状态,并记录为 Committed OpenSpec。
- Phase 3 所需的 proposal、design、specs 和 tasks 均存在且一致。
- 所有 Phase 3 preflight 风险已消除或明确记录为已接受。
**输出**:
- Committed OpenSpec 状态说明。
- Phase 3 preflight 检查结果,默认写入 `decisions.md` 或 `acceptance.md`。
**Human checkpoint**:
- 用不超过五句话说明 Committed OpenSpec 的范围、接口影响、剩余风险和执行计划。
- 询问是否进入 Phase 3 OpenSpec apply;除非用户在启动时明确要求“全自动执行”,必须等待用户明确说出进入 apply、开始实现、执行修改、继续 Phase 3 或等价授权。
- 不得把 Phase 2 的单个 grill 决策确认当作本 checkpoint 的授权。
## Phase 3 — OpenSpec apply
**进入条件**:
- `openspec/changes/<slug>/` 中 proposal/design/specs/tasks 已达到可执行状态。
- Phase 2.5 后已获得用户继续实现的确认,除非用户要求“全自动执行”。
- `openspec/changes/<slug>/` 中 proposal/design/specs/tasks 已通过 Phase 2.9,成为 Committed OpenSpec。
- Phase 2.9 后已获得用户明确的 Phase 3 apply 授权,除非用户在启动时要求“全自动执行”。
- devflow 与 OpenSpec 没有未解决冲突。
- 没有未解决的 user-interview 问题、未判级接口影响、未汇报 evidence-driven 结论或未接受架构风险。
**显式子 skill**:`openspec-apply-change`;遇到 bug/不确定行为时显式调用 `diagnose`;需要测试驱动时显式调用 `tdd`。进入本阶段必须先声明调用方式,不能静默 fallback。
@@ -163,6 +209,12 @@
- 优先调用 `openspec-apply-change`。
- 执行依据是 OpenSpec specs/tasks;devflow 只能作为上下文参考。
- 按 OpenSpec tasks 的纵向切片实现。
- 当用户质疑、用户要求修改、代码检查、测试失败或运行行为与 OpenSpec 冲突时,先分类再继续:
- 实现偏差:Committed OpenSpec 正确,代码没有按规格做;修代码,不改 OpenSpec。
- 规格遗漏:OpenSpec 没覆盖真实边界、接口影响、验收口径或外部可见行为;暂停 apply,修正 OpenSpec 后重新提交。
- 设计冲突:OpenSpec 与架构、ADR、历史验收、数据所有权或模块边界冲突;暂停并向用户确认设计方向。
- 用户变更:用户改变目标、范围、优先级、验收或风险接受度;更新 proposal/specs/tasks 后再继续。
- 冲突分类、证据、用户确认和 OpenSpec 回写状态必须记录到 `decisions.md` 或 `acceptance.md`。
- 当用户要求、行为复杂或回归风险高时使用 TDD。
- 当测试失败、行为意外或原因不确定时使用 diagnose。
- 如果 diagnose 发现根因是 OpenSpec 不准确,先修正 OpenSpec,再继续 apply。
@@ -170,6 +222,7 @@
**退出条件**:
- OpenSpec tasks 已完成,或剩余 tasks 已明确记录。
- 所有实现期冲突已分类并处理;没有未确认的规格遗漏、设计冲突或用户变更。
- 已运行验证,或记录了未验证原因。
- 已列出已知限制。
@@ -188,10 +241,12 @@
- 从 OpenSpec、实现结果和验证结果提炼 brief、evidence、decisions 和 acceptance;只在复杂场景按需拆出 PRD/research/design/tasks/alignment。
- 写入或更新验收记录,并区分静态验证、脚本验证、浏览器/人工验证、未验证。
- 如果本次流程产生可复用经验,写入 compound knowledge。
- 更新 `devflow/index.md`,记录日期、slug、领域、关键词、关联 OpenSpec 和状态。
- 询问用户是否要 archive OpenSpec change;不要默认执行归档。
**退出条件**:
- `devflow/projects/YYYY-MM-DD-{slug}/` 能说明做了什么、为什么做、OpenSpec 如何指导执行、还剩什么、如何验证。
- `devflow/index.md` 已包含或更新本项目条目。
- 用户已被询问是否 archive OpenSpec change。
**输出**:
@@ -65,6 +65,68 @@
- 风险接受:{accepted by whom/when}
```
## 接口影响记录模板
```markdown
# {标题} 接口影响记录
## 分级
- 级别:L1 内部实现 / L2 内部接口 / L3 协作接口 / L4 破坏性接口
- 判级原因:{why this level}
- 是否需要独立接口文档:是 / 否
## 变更对象
- 接口/字段/DTO/事件/回调/数据库契约:
- 判断逻辑变化:
- 可观察行为变化:返回数据 / 状态 / 错误码 / 权限结果 / 过滤排序 / 幂等性 / 时序 / 副作用 / 无
## 影响范围
- 调用方/消费者:
- 是否跨模块/跨服务/跨团队:
- 旧调用方是否需要改动:
## 兼容与迁移
- 是否向后兼容:
- 迁移/灰度/回滚要求:
- 风险接受:
## 验收方式
- 如何证明新行为正确:
- 如何证明旧行为未破坏:
- 需要用户确认的问题:
```
## 实现期冲突记录模板
```markdown
# {标题} 实现期冲突记录
## 冲突摘要
- 触发来源:用户质疑 / 用户变更 / 代码发现 / 测试失败 / 运行行为
- 冲突对象:proposal / design / specs / tasks / ADR / 代码行为
- 分类:实现偏差 / 规格遗漏 / 设计冲突 / 用户变更
## 证据
- OpenSpec 依据:
- 代码或测试证据:
- 用户反馈:
## 处理
- 决策:
- 是否需要用户确认:是 / 否
- OpenSpec 回写:不需要 / 已回写 / 待回写
- 代码处理:
- 验证方式:
```
## PRD 模板
```markdown
-259
View File
@@ -1,259 +0,0 @@
---
name: essence
description: Invoke when a project is too large or you only want the core design insights. Extracts 1-2 standout design patterns with deep analysis, lens-guided perspectives, and migration examples. Not for full project analysis or quick lookups.
metadata:
version: "0.5.0"
---
# Essence: Extract Core Design Patterns
Prefix your first line with 🥷 inline, not as its own paragraph.
You are a jewel inspector. A project has thousands of files — your job is to find the one or two brilliant ideas worth stealing.
**This is NOT a lite version of `/explore`.** `/explore` reads the whole project and summarizes at the end. `/essence` goes deep on one thing and ignores everything else.
## Mode Selection
First, check whether an `/explore` result exists:
- `/explore` report exists → it already identified 2-3 core designs, default to **User-directed**. Ask the user which design to deep-dive, or whether to switch mode.
- No `/explore` result → this is an independent launch, default to **Auto-detect**.
Always confirm before proceeding:
| Mode | When | Entry |
|---|---|---|
| **User-directed** | Already have a design target from `/explore`, or know exactly which design to investigate | User tells you what to look for |
| **Auto-detect** | Independent launch, project is large, want the AI to find the standout design | You find the standout design |
| **Lens-guided** | "Analyze this from a [mechanical/intentional/evolution] perspective" | Apply a specific analytical lens |
### Lens definitions
| Lens | Core question | Guided behavior |
|---|---|---|
| **Mechanical** (default) | How does it work? | Read source code, trace call chains, examine interfaces |
| **Intentional** | Why this way? | Read design docs/RFCs/PRs, extract decision rationale and tradeoffs |
| **Evolution** | How did it get here? | Read git history/changelog, compare before/after, identify migration drivers |
A lens shapes which sources to read and how to frame the output, but does not add separate phases.
### Auto-detect signals
A design is "essence" if it passes 2 or more of these signals:
| Signal | Evidence |
|---|---|
| README highlights it prominently | "Built on a plugin architecture" as a headline feature |
| Has standalone architecture docs | ARCHITECTURE.md, docs/design/, blog post by author |
| Heavily discussed in Issues/PRs | Design decisions debated by community |
| Unique among similar projects | Competitors don't do it this way |
| Rich design comments in code | JSDoc/TSDoc explaining why, not what |
| Cross-module contract | A type, interface, or protocol imported across module boundaries (not just files). Go: most-implemented interface. Python: most-subclassed abstract base. Rust: most-implemented trait. These define subsystem relationships. |
| File size anomaly | One file is disproportionately large or small for its responsibility — signals non-trivial logic |
| Dedicated test coverage | Tests specifically validate this design's behavior, not just happy paths |
**"Clean code" is NOT a signal.** A well-written utility function is not essence. An architecture decision that shapes the entire project is.
If no design passes 2+ signals, tell the user: "This project has no standout design. Try `/explore` for a full analysis instead."
## Phase 1: Locate
**User-directed mode:**
- Go directly to the directory or file the user names.
- If the directory doesn't exist, stop and tell the user. Do NOT invent an alternative.
**Auto-detect mode:**
- Scan README, CLAUDE.md, and top-level docs for architecture claims.
- Identify 1-2 standout design directions.
- Present to the user: "The standout designs appear to be: A) {design A}, B) {design B}. Which should we dive into?"
- If user doesn't choose, pick the strongest one and state why.
**Lens-guided mode:**
- Confirm the lens with the user (Mechanical/Intentional/Evolution).
- Frame the search in terms of the lens.
- Example: "You want the Mechanical view — I'll trace the core implementation and extract the pattern."
**Output:** 1-2 design directions to analyze + lens confirmation.
**Stall signal:** Cannot identify any standout design → the project may be a conventional CRUD app or wrapper. Stop and recommend `/explore` or a different project.
## Phase 2: Deep Dive
Read the core files related to the chosen design. Maximum 10 files. Let the lens guide source selection: Mechanical → source code and type definitions; Intentional → design docs, RFCs, PR discussions; Evolution → git history, changelog, migration guides.
**For each file:**
- What role does it play in this design?
- What interfaces does it expose?
- How does it connect to other parts of the system?
**Trace the call chain:**
- Start from the entry point that uses this design.
- Follow the flow until you understand the full pattern.
- Stop when you hit boilerplate, config, or test files.
**Output:** Core file list (≤10) + call chain + lens-specific annotations.
**Stall signal:** The design spans more than 10 files and you can't find the boundary → the design is probably the project's core architecture. Switch to `/explore` for a full analysis instead.
## Phase 3: Extract Pattern
Analyze the design at a higher level. Let the lens shape the analysis angle:
- **Mechanical** → emphasize structure, interfaces, data flow — produce a pattern diagram + interface contracts
- **Intentional** → emphasize decision rationale, tradeoffs — produce a decision record (context → options → rationale)
- **Evolution** → emphasize before/after comparison, migration drivers — produce a timeline + catalyst events
**Universal analysis dimensions** (all lenses):
- **Problem:** What specific problem does this design solve? What was the pain before?
- **Pattern:** What's the name of this pattern? (Named: MVC, Observer, Plugin, Middleware. Custom: describe it in one sentence.)
- **Alternatives:** What simpler or more complex approaches could solve the same problem?
- **Tradeoffs:** Why did the author choose this? What does it give up?
- **Evidence:** What in the code proves this analysis is correct? (Specific files, functions, comments.)
**Output:** Design pattern card (lens-framed).
**Stall signal:** Cannot explain why the author chose this design over alternatives → read commit messages and PR discussions for design rationale. If unavailable, state "author's reasoning unknown" in the report.
## Phase 4: Migrate
Make the learning actionable. Let the lens tailor the output:
- **Mechanical** → copy-paste code skeleton (≤20 lines with TODOs)
- **Intentional** → decision framework (checklist for evaluating tradeoffs)
- **Evolution** → migration path (step-by-step refactor plan)
**Universal deliverables** (all lenses):
- **Can you use this?** Is the design applicable to the user's own projects? If not, why?
- **Steal-it example:** A simplified version (under 20 lines) that captures the core idea. Not production code — a teaching example.
- **Pitfalls:** What context does this design depend on? What would break if you copy it blindly?
**Output:** Migration example + pitfall list (lens-tailored).
**Stall signal:** The design depends on framework internals, language features, or ecosystem the user doesn't have → explain the core idea abstractly instead of providing code.
## Phase 5: Self-review
Check the report is honest:
**All modes:**
- [ ] The design is real (not inferred, not imagined). Evidence: specific files cited.
- [ ] The analysis is deep enough that you could explain it out loud.
- [ ] The migration example captures the core idea, not surface syntax.
- [ ] Pitfalls are specific, not vague ("needs X version" not "may not work everywhere").
**Stall signals (any one → return to relevant phase):**
- Cannot name a file that proves the pattern → back to Phase 2
- Cannot explain why it's better than alternatives → back to Phase 3
- Migration example is over 20 lines → simplify, back to Phase 4
- Lens-specific check failed (e.g., Mechanical missing end-to-end call chain, Intentional missing decision rationale, Evolution missing timeline) → back to relevant phase
**Output:** Essence report with lens annotation.
## Optional: HTML Card
**Only when the user explicitly requests it.**
Generate an HTML visualization card as a shareable deliverable.
### HTML Card Structure (Glassmorphism 2.0 - Essence Variant)
```html
<!DOCTYPE html>
<html lang="zh-CN">
<head>
<meta charset="UTF-8">
<title>{Project Name} - Essence Report</title>
<script src="https://cdn.tailwindcss.com"></script>
<script src="https://cdn.jsdelivr.net/npm/mermaid/dist/mermaid.min.js"></script>
<style>
/* Same glassmorphism styles as /explore */
:root { --glass-bg: rgba(255,255,255,0.4); --primary: #8b5cf6; }
[data-theme="dark"] { --glass-bg: rgba(15,23,42,0.6); --primary: #a78bfa; }
.glass-panel { backdrop-filter: blur(12px); border-radius: 1rem; }
.pattern-diagram { font-family: monospace; background: rgba(0,0,0,0.03); }
</style>
</head>
<body class="p-8">
<nav class="fixed top-4 left-1/2 -translate-x-1/2 w-[90%] max-w-4xl glass-panel z-50 px-6 py-3">
<span class="font-bold text-xl">💎 {Project Name} 精华</span>
<span class="text-sm opacity-70">Lens: {lens} | Pattern: {pattern_name}</span>
</nav>
<main class="max-w-4xl mx-auto mt-24 space-y-6">
<section class="glass-panel p-6">
<h2 class="text-xl font-bold mb-4">🎯 Design Analyzed</h2>
<p>{one-line description}</p>
</section>
<section class="glass-panel p-6">
<h2 class="text-xl font-bold mb-4">🔷 Pattern ({lens})</h2>
<!-- Lens-framed pattern card -->
</section>
<section class="glass-panel p-6">
<h2 class="text-xl font-bold mb-4">🔗 Call Chain</h2>
<pre class="mermaid">{diagram}</pre>
</section>
<section class="glass-panel p-6">
<h2 class="text-xl font-bold mb-4">📦 Migration Example</h2>
<pre class="pattern-diagram"><code>{code_example}</code></pre>
<p class="text-sm opacity-70 mt-2">Pitfalls: {pitfalls}</p>
</section>
</main>
<script>mermaid.initialize({ startOnLoad: true });</script>
</body>
</html>
```
### Output Format
```markdown
### HTML Card Generated
- **Path:** `outputs/{project}-essence.html`
- **Theme:** {modern/ink}
- **Accent Color:** Purple (essence = jewel)
```
**When to skip:** Skip HTML generation unless the user requests it or the analysis is production-critical. When HTML generation fails, deliver a plain-text report instead.
---
## Hard Rules
- **No code evidence = no conclusion.** Every claim about a design must cite a specific file, function, or comment.
- **Under 20 lines for migration examples.** If you can't explain the idea in 20 lines, you don't understand it well enough.
- **Stop after the report.** Do not modify the user's project or the target project.
- **HTML is optional.** Do not block analysis on HTML generation.
## Gotchas
| What happened | Rule |
|---|---|
| 提取的"精华"是 AI 脑补的 | 必须有代码证据(文件 + 行号),不写空泛结论 |
| 用户指定方向但该模块不存在 | 停止并告知用户,不编造替代方向 |
| 项目没有 standout 设计(胶水代码) | 标记"无可提取精华",建议改用 `/explore` |
| Phase 4 迁移示例超过 20 行 | 简化到核心思路,不是复制生产代码 |
| 分析了一个小工具函数 | 工具函数不是设计。设计影响整个架构,工具只解决一个问题 |
| 从 commit message 推断作者意图但没有代码佐证 | Commit message 是辅助证据,必须有代码结构本身的支持 |
| 透镜模式选错导致输出不符预期 | Phase 1 先确认透镜,Mechanical 读代码、Intentional 读文档、Evolution 读历史 |
| 透镜分析流于表面 | 每个透镜有特定输出格式:Mechanical→图 + 接口,Intentional→决策记录,Evolution→时间线 |
| HTML 卡片生成失败 | 降级到纯文本报告,不阻塞分析交付 |
## Outcome
```
Essence Report: {project name}
Lens: mechanical / intentional / evolution
Design analyzed: {one-line description}
Files examined: {count}
Pattern: {pattern name or custom description}
Migration: {steal-it example, ≤20 lines}
HTML generated: yes / no
Status: complete
```
After the report, stop. No modifications. No follow-ups.
@@ -1,79 +0,0 @@
# Essence Detection Signals
How to identify the standout design in a project when the user doesn't specify a direction.
## Signal Strength
A design passes the "essence" threshold if it scores 2+ signals.
### Strong Signals (score = 1 each)
| Signal | How to detect | Example |
|---|---|---|
| **README headline** | Project name is followed by a design claim | "Vite — Next generation frontend tooling with **ESM-first architecture**" |
| **Architecture docs** | Standalone design document exists | `ARCHITECTURE.md`, `docs/design/`, `docs/architecture/` |
| **Official blog post** | Author wrote about the design on their blog | tw93.fun, Vite blog, React blog posts |
| **Community discussion** | Issues/PRs debate the design decision | "Why we chose X over Y" discussions with many comments |
| **Rich code comments** | JSDoc/TSDoc explaining WHY, not WHAT | "We use this pattern because..." with detailed reasoning |
### Objective Signals (score = 1 each, no subjective judgment needed)
| Signal | How to detect | Example |
|---|---|---|
| **Cross-module contract** | A type, interface, or protocol imported across module boundaries (not just files). Go: most-implemented interface. Python: most-subclassed abstract base. Rust: most-implemented trait. | `Plugin` interface implemented by 8 subsystems, each in its own package |
| **File size anomaly** | One file's line count is ≥3× the median for its category (handlers, utils, etc.) | Average handler: 50 lines. One handler: 800 lines with state machine logic |
| **Dedicated test coverage** | Tests exist specifically for this design's edge cases, not just happy paths | `plugin.test.ts` tests plugin resolution, fallback, lifecycle — not just "it loads" |
### Weak Signals (score = 0.5 each)
| Signal | How to detect | Example |
|---|---|---|
| **Unique among competitors** | Same category, different architecture | Next.js uses SSR, Remix uses nested routes — that difference IS the essence |
| **Most-starred files** | GitHub shows stars/bookmarks on specific files | "This file has 200+ stars on GitHub" |
| **Core algorithm** | One file contains non-trivial logic that drives the project | Diff algorithm, compiler pass, state machine |
| **API design** | The public API is notably elegant or unusual | `create()` returns a builder chain, not an object |
## Not Signals
These do NOT count as essence:
- "Clean code" or "well organized" — that's quality, not design
- "Uses TypeScript" — that's a language choice, not architecture
- "Has good tests" — that's engineering discipline, not design
- "Many stars on the repo" — popularity ≠ design quality
- "Uses the latest framework" — following trends ≠ standing out
- Utility functions — even well-written ones are tools, not designs
## Auto-detect Procedure
When the user says "find the essence":
1. **Read README fully.** What is the #1 feature the author leads with? That's a candidate.
2. **Check for design docs.** Is there `ARCHITECTURE.md` or equivalent? That's a candidate.
3. **Scan the import graph.** Which file is imported by the most other files? Use `grep -r "import.*from" src/ | sort | uniq -c | sort -rn` or equivalent. The top result is likely the core.
4. **Check file sizes.** Are any files disproportionately large or small for their apparent role? That signals hidden complexity.
5. **Check uniqueness.** Compare with 1-2 well-known alternatives. What does this project do differently?
6. **Present 1-2 candidates** to the user with evidence. Let them choose or auto-select the strongest.
### Example Output Format
```
Standout designs in {project}:
A) {Design A name} — evidenced by {README claim / file / doc}
What it does: {one sentence}
B) {Design B name} — evidenced by {code comment / unique feature / community discussion}
What it does: {one sentence}
Which should we dive into? (or I can pick the strongest)
```
## Failure Modes
| Situation | Response |
|---|---|
| No signal passes 2+ threshold | "This project uses conventional architecture. Try `/explore` for a full analysis, or pick a more architecturally interesting project." |
| User-specified module doesn't exist | Stop. Do NOT suggest an alternative. Tell the user the path doesn't exist. |
| Project is a wrapper (thin layer over another tool) | "This project is primarily a wrapper around {X}. The design is in {X}, not here. Try analyzing {X} instead." |
| Project is configuration-only (just JSON/YAML files) | "This project has no code architecture. It's configuration-driven. Try `/explore` for a full overview instead." |
-87
View File
@@ -1,87 +0,0 @@
---
name: explore
description: Invoke when you need project-level understanding and an onboarding path. Produces a project learning report for code and non-code repositories with fixed phases for positioning, structure, flow, start path, and core designs. Not for deep code extraction or interactive teaching.
metadata:
version: "0.5.0"
---
# Explore: Project Understanding and Onboarding
Prefix your first line with 🥷 inline, not as its own paragraph.
You are a project cartographer. Your job is to help the user understand what a project is, why it is worth studying, how it is organized, and where to start.
`/explore` is the entry point for first contact with a repository or project-like artifact. It builds global understanding. It does not perform code-level essence extraction and it does not run interactive teaching.
## Project Type Detection
After the initial scan, classify the target before continuing:
| Type | Signals | What changes |
|---|---|---|
| **Code repository** | `go.mod`, `pyproject.toml`, `Cargo.toml`, source directories, executable entrypoints | Run all 4 phases |
| **Skill / docs / knowledge repository** | `SKILL.md`, mostly Markdown, docs-first structure, no runnable application entrypoint | Skip Phase 2 (Flow) and Phase 3 (Start Path) |
| **Template / scaffold repository** | Starter files, minimal logic, setup-first repo | Phase 2 may stay structural and Phase 3 may be minimal |
State the detected type before proceeding. If uncertain, say what evidence is missing and continue with the closest matching type.
## Phase 1: Positioning & Structure
- What this project is, why it is worth studying, and who it is for.
- Top-level structure: main modules, documents, directories, and the likely learning entry area.
- Tradeoffs vs alternatives when evidence exists.
## Phase 2: Flow
**Code repositories only.**
- Skip for non-code and template repositories.
- Trace the main runtime or request flow.
- Produce at least one architecture or core-flow diagram.
- Keep the trace focused on the golden path rather than exhaustive coverage.
## Phase 3: Start Path
**Code repositories only when runnable or meaningfully inspectable.**
- Provide the minimal path to start learning or running the project.
- Give the first command or first inspection step.
- Suggest one safe first modification or observation point when appropriate.
## Phase 4: Core Designs
- Summarize 2-3 core implementations or ideas.
- Keep this at overview depth.
- For each item, include what it is, where it lives, and why it matters.
## Minimum Deliverables
The final `/explore` report must include:
- Project positioning
- Why it is worth studying
- 2-3 core implementations or core ideas
- Tradeoffs or comparisons when applicable
- At least 1 diagram:
- code repository → architecture diagram or core flow diagram
- non-code repository → structure diagram, idea map, or workflow diagram
## Boundary Rules
`/explore` may:
- scan structure
- explain the main flow
- provide a minimal start path
- summarize 2-3 core designs
`/explore` must not:
- perform `/essence`-level deep extraction
- act as `/follow`-style guided teaching
- include Verify, Deep Fission, or HTML Output phases
- preserve no retired lightweight fallback behavior
## Outcome
```
Explore Report: {project name}
Project type: code / skill-docs / template
Phases completed: 4/4 (or note skipped code-only phases)
Diagram included: yes / no
Core designs: 2-3
Status: complete
```
After the report, stop. Do not proceed to `/essence` or `/follow` automatically.
@@ -1,98 +0,0 @@
# Project Analysis Methods
How to read and understand an unfamiliar code project.
## 1. Identify the Entry Point
Every project has a door. Find it first.
### By Language
| Language | Look for |
|---|---|
| **JavaScript/TypeScript** | `package.json` → `main` / `bin` / `scripts.dev` |
| **Python** | `setup.py` → `entry_points`, `pyproject.toml` → `[project.scripts]`, or top-level `app.py` / `main.py` / `__main__.py` |
| **Go** | `package main` in any file, conventionally `main.go` or `cmd/*/main.go` |
| **Rust** | `src/main.rs` or `src/bin/*.rs` |
| **Java** | Class with `public static void main(String[] args)` |
| **C/C++** | `main()` function, conventionally in `src/main.c` |
| **Swift** | `main.swift` or file with `@main` attribute |
### In Frameworks
| Framework | Entry point |
|---|---|
| Next.js | `app/` or `pages/` directory, `next.config.js` |
| React (Vite) | `src/main.tsx` or `src/main.jsx` |
| Vue (Vite) | `src/main.ts` or `src/main.js` |
| Express | File that calls `app.listen()` |
| FastAPI | File that creates `FastAPI()` instance |
| Django | `manage.py`, then project name directory with `urls.py` / `wsgi.py` |
| Flask | `app.py` or `app/__init__.py` |
| Spring Boot | `*Application.java` with `@SpringBootApplication` |
## 2. Judge Project Complexity
Don't over-engineer simple projects. Don't under-analyze complex ones.
### Simple (<50 files, single language)
- Read every source file.
- No need for flow diagrams beyond a simple sequence.
- A light `/explore` pass is probably enough.
### Standard (50-500 files, 1-2 languages)
- Read entry point + core modules + 1-2 feature files.
- Build 1-2 flow diagrams.
- `/explore` is the right level.
### Complex (>500 files, multi-language, monorepo)
- Read entry point + architecture docs + one representative module.
- Use `/essence` to find standout designs, or `/explore` for one package at a time.
- Do NOT try to understand the whole project in one pass.
## 3. Separate Core Code from Scaffolding
Not all files are worth reading.
### Ignore (scaffolding)
- `*.config.js`, `*.config.ts` — configuration, not logic
- `dist/`, `build/`, `out/` — generated output
- `node_modules/`, `vendor/`, `.venv/` — dependencies
- `*.lock`, `yarn.lock`, `go.sum` — lock files
- `LICENSE`, `CODEOWNERS`, `.editorconfig` — project meta
- `test/fixtures/`, `test/data/` — test data
### Read (core)
- Entry point file
- Router/middleware/config handlers
- Model/entity/schema definitions
- Core algorithm or business logic files
- Files referenced most in imports
### Hint: Follow imports
```
entry file → import A → import B → core logic
```
Each import is a dependency. Follow the chain until you hit a file that doesn't import anything else — that's usually the core.
## 4. Read Unfamiliar Framework Code
You don't know every framework. That's fine.
### Strategy
1. **Find the routing layer first.** Every framework has a way to map URLs or events to handlers. Find it. It tells you the project's capabilities.
2. **Follow ONE request end-to-end.** Don't try to understand all routes. Pick the simplest one (often "health check" or "get by ID") and trace it from entry to response.
3. **Identify the framework's conventions.** Most frameworks follow a pattern:
- MVC: Controller → Model → View
- Middleware: Request → Middleware chain → Handler → Response
- Component: Parent renders children, props flow down, events flow up
- Plugin: Core calls hooks, plugins register handlers
4. **Don't fight the framework's abstraction.** If the project uses ORM, don't look for raw SQL. If it uses dependency injection, don't look for `new()` calls. Understand what abstraction layer they chose.
5. **Use the framework's own docs.** If stuck on "how does this framework work?", check the official docs. Don't reverse-engineer what's documented.
-173
View File
@@ -1,173 +0,0 @@
# Flow Pattern Library
Common architecture patterns and how to identify them in code.
## MVC / MVVM / MVX
### What it is
Separation of data (Model), UI/presentation (View), and coordination logic (Controller/ViewModel).
### File signatures
| Pattern | Directories/Files |
|---|---|
| **MVC** | `controllers/`, `models/`, `views/` |
| **MVVM** | `viewmodels/`, `views/`, `models/` |
| **Layered** | `app/`, `domain/`, `infrastructure/` (Clean/Hexagonal) |
### Flow
```
Request → Controller → Model (data) → View (render) → Response
```
### Key question
"Does the file handle data, display, or coordination?" If yes → MVC-family.
---
## Middleware Chain
### What it is
Each handler processes the request and passes it to the next. Like an assembly line.
### File signatures
| Framework | Indicator |
|---|---|---|
| **Express/Koa** | `app.use(...)`, `app.get('/', handler)` |
| **FastAPI** | `@app.middleware("http")`, `Depends()` |
| **Next.js** | `middleware.ts` at root or in `app/` |
| **Gin (Go)** | `router.Use(middleware1, middleware2)` |
| **Koa** | `app.use(async (ctx, next) => { ... })` |
### Flow
```
Request → Middleware A → Middleware B → Handler → Response
↓ ↓
auth check log request
```
### Key question
"Does this function call `next()` or pass control to something else?" If yes → middleware.
### Common middleware order
```
1. CORS / Security headers
2. Logging / Request ID
3. Authentication / Authorization
4. Body parsing / Validation
5. Rate limiting
6. Route handler
7. Error handler (catches everything above)
```
---
## Plugin / Extension System
### What it is
Core provides hooks or interfaces. External code registers handlers. The core doesn't know about specific plugins.
### File signatures
| Pattern | Indicator |
|---|---|
| **Hook-based** | `registerHook('eventName', handler)`, `hooks.on('event', fn)` |
| **Interface-based** | Abstract class or interface that plugins implement |
| **Discovery-based** | Directory scan (`plugins/`), import all, register by convention |
| **VSCode-style** | `contributes` in `package.json`, activation events |
### Flow
```
Core starts
↓
Scans for plugins
↓
Each plugin registers itself
↓
Core fires hooks → plugins respond
↓
Core runs with extended capabilities
```
### Key question
"Can I add functionality without modifying core code?" If yes → plugin architecture.
---
## Event-Driven
### What it is
Components communicate through events, not direct calls. Publishers emit, subscribers listen.
### File signatures
| Pattern | Indicator |
|---|---|
| **Node EventEmitter** | `eventEmitter.on('event', handler)`, `eventEmitter.emit('event', data)` |
| **Pub/Sub** | `pubsub.subscribe('channel', handler)`, `pubsub.publish('channel', data)` |
| **Redux-style** | `dispatch(action)`, `reducer(state, action) → newState` |
| **Observable** | `observable.subscribe(fn)`, `pipe(map, filter)` |
| **Signals (Python)** | `@signal.connect`, `signal.send()` |
### Flow
```
Component A emits "user.created"
↓
Listener B hears it → sends welcome email
Listener C hears it → creates default settings
Listener D hears it → logs analytics
```
### Key question
"Does code communicate without importing or calling each other directly?" If yes → event-driven.
---
## State Management
### What it is
Centralized storage for application state. Components read and update through defined interfaces.
### File signatures
| Pattern | Indicator |
|---|---|
| **Redux** | `createStore()`, `dispatch()`, `useSelector()`, `@reduxjs/toolkit` |
| **Zustand** | `create((set) => ({ ... }))` |
| **Jotai** | `atom(value)`, `useAtom(atom)` |
| **MobX** | `@observable`, `@action`, `@computed` |
| **React Context** | `createContext()`, `useContext()`, `Provider` |
| **Pinia (Vue)** | `defineStore()`, `state`, `actions` |
### Flow
```
Component dispatches action
↓
Reducer processes action + current state
↓
New state emitted
↓
Subscribed components re-render
```
### Key question
"Where does the app store data that multiple components need?" If it's a single store → state management pattern.
---
## Pipeline / Chain of Responsibility
### What it is
Data flows through a series of processors. Each processor transforms the data and passes it on.
### File signatures
| Pattern | Indicator |
|---|---|
| **Stream processing** | `.pipe(transform1).pipe(transform2)` |
| **Compiler/lexer** | Source → Tokenize → Parse → Transform → Generate |
| **Data pipeline** | `input → transform → validate → output` |
| **Makefile** | Target depends on prerequisites, each is a step |
### Flow
```
Raw input → Tokenizer → Parser → Transformer → Generator → Output
```
### Key question
"Does data get progressively transformed through a fixed sequence of steps?" If yes → pipeline.
-120
View File
@@ -1,120 +0,0 @@
#!/usr/bin/env bash
# Collect project structure for /explore analysis.
# Usage: Run from project root, or pass project path as argument.
# Output: Structured text with directory tree, file counts, language distribution.
set -euo pipefail
PROJECT_DIR="${1:-.}"
cd "$PROJECT_DIR"
echo "=== PROJECT STRUCTURE ==="
echo ""
# Directory tree (depth 3, exclude common noise)
echo "--- Directory Tree (depth 3) ---"
if command -v tree &>/dev/null; then
tree -L 3 \
-I "node_modules|vendor|.git|dist|build|out|.venv|__pycache__|*.egg-info|coverage|.nyc_output" \
--dirsfirst
elif command -v find &>/dev/null; then
find . -maxdepth 3 \
-not -path "./.git/*" \
-not -path "./node_modules/*" \
-not -path "./vendor/*" \
-not -path "./dist/*" \
-not -path "./build/*" \
-not -path "./out/*" \
-not -path "./.venv/*" \
-not -path "*/__pycache__/*" \
-not -path "*/.egg-info/*" \
-not -path "*/coverage/*" \
-not -path "./.nyc_output/*" \
-print | head -100 | sort
fi
echo ""
echo "=== FILE COUNTS ==="
echo ""
# Count files by extension (top 10)
echo "--- Top 10 File Types ---"
find . -type f \
-not -path "./.git/*" \
-not -path "./node_modules/*" \
-not -path "./vendor/*" \
-not -path "./dist/*" \
-not -path "./build/*" \
-not -path "./out/*" \
-not -path "./.venv/*" \
-not -path "*/__pycache__/*" \
-printf '%f\n' | \
sed 's/.*\.//' | \
grep -v '^\.[^/]*$' | \
sort | uniq -c | sort -rn | head -10
echo ""
echo "=== TOTAL FILE COUNT ==="
echo ""
# Total files (excluding noise)
total=$(find . -type f \
-not -path "./.git/*" \
-not -path "./node_modules/*" \
-not -path "./vendor/*" \
-not -path "./dist/*" \
-not -path "./build/*" \
-not -path "./out/*" \
-not -path "./.venv/*" \
| wc -l)
echo "Total source files: $total"
echo ""
echo "=== DEPENDENCY FILES ==="
echo ""
# List dependency declaration files found
for dep_file in "package.json" "requirements.txt" "pyproject.toml" "setup.py" "go.mod" "go.sum" "Cargo.toml" "Cargo.lock" "pom.xml" "build.gradle" "Gemfile" "Gemfile.lock" "composer.json"; do
if [ -f "$dep_file" ]; then
echo "FOUND: $dep_file"
fi
done
# Check for workspace/monorepo configs
echo ""
echo "=== WORKSPACE / MONOREPO ==="
echo ""
for ws_file in "turbo.json" "nx.json" "lerna.json" "pnpm-workspace.yaml" "go.work"; do
if [ -f "$ws_file" ]; then
echo "FOUND: $ws_file"
fi
done
# Check Cargo.toml for workspace
if [ -f "Cargo.toml" ] && grep -q '\[workspace\]' Cargo.toml 2>/dev/null; then
echo "FOUND: Cargo.toml [workspace]"
fi
echo ""
echo "=== ENTRY POINTS ==="
echo ""
# Try to identify entry points
if [ -f "package.json" ]; then
main=$(node -e "try{const p=require('./package.json');console.log(p.main||'');}catch(e){}" 2>/dev/null || echo "")
bin=$(node -e "try{const p=require('./package.json');console.log(typeof p.bin==='string'?p.bin:JSON.stringify(p.bin));}catch(e){}" 2>/dev/null || echo "")
dev=$(node -e "try{const p=require('./package.json');console.log(p.scripts?.dev||p.scripts?.start||'');}catch(e){}" 2>/dev/null || echo "")
[ -n "$main" ] && echo "package.json main: $main"
[ -n "$bin" ] && echo "package.json bin: $bin"
[ -n "$dev" ] && echo "package.json dev/start: $dev"
fi
for entry in "src/main.ts" "src/main.tsx" "src/main.js" "src/main.jsx" "src/index.ts" "src/index.js" "src/main.py" "app/main.py" "main.go" "src/main.rs" "app.py" "index.js" "index.ts"; do
if [ -f "$entry" ]; then
echo "FOUND: $entry"
fi
done
echo ""
echo "=== COLLECTED ==="
-101
View File
@@ -1,101 +0,0 @@
---
name: follow
description: Invoke when the user wants an interactive learning session based on an existing `/explore` or `/essence` report. Guides runnable or reader-style follow-along sessions. Not for fresh project analysis or pattern-only extraction.
metadata:
version: "0.5.0"
---
# Follow: Guided Learning Session
Prefix your first line with 🥷 inline, not as its own paragraph.
You are a guide. The user wants to learn from a project step by step with help, context, and correction. You guide the learning process, but you do not replace it.
`/follow` is not a fresh project analyzer. It only works from an existing `/explore` or `/essence` result.
## Pre-check
`/follow` only works when there is already an `/explore` report or an `/essence` report.
- `/explore` report exists → use it as the main learning path
- `/essence` report exists → use it for design-focused guided study
- Neither exists → refuse clearly
Refusal behavior:
"I need an existing `/explore` or `/essence` result before I can guide a follow-along session. Please run `/explore` for project understanding or `/essence` for a focused deep dive first."
Load the existing report before continuing.
## Mode Selection
After the pre-check, select one mode based on the prerequisite report:
- From `/explore` + code repository → default **Runnable**
- From `/explore` + non-code repository → force **Reader**
- From `/essence` → default **Reader** (user is in design-analysis state)
| Mode | When | Entry |
|---|---|---|
| **Runnable** | Report confirms the project is a runnable code repository and the user wants to learn by running and changing it | Start from environment and first execution |
| **Reader** | Project has no runtime, or the user is studying design/architecture, or the prerequisite report is from `/essence` | Start from guided reading |
State the selected mode before proceeding. Do not re-scan the project — use the prerequisite report to decide.
## Teaching Interaction Rules
`/follow` must teach by guidance, not by dumping answers:
- explain the purpose of the current step first
- give the user an observation point or action point
- ask the user to predict, try, or explain before revealing the answer
- then reveal, correct, or deepen the explanation
- never say "go read the code" as a standalone instruction. When referencing code, always start with: what design idea this code embodies, why it matters in the overall architecture, and what the user should pay attention to
## Runnable Check
Before Runnable mode, confirm from the **prerequisite report** (do not re-scan the project):
- If the report identified the target as a code repository with a recognized runtime (`go.mod`, `pyproject.toml`, `Cargo.toml`, `Makefile`, `build.gradle`, `pom.xml`, `CMakeLists.txt`, etc.), proceed with Runnable.
- If the report classified it as non-code, or no runtime entrypoint was found, switch to Reader and explain why.
- If the prerequisite is `/essence`, confirm with the user: essence is design-focused, Reader is the natural fit. Allow Runnable only if the user explicitly insists.
- Do not introduce a third mode.
## Runnable Mode Flow
1. Confirm environment and prerequisites.
2. Let the user run the project.
3. Let the user make one safe change.
4. Walk the main flow together.
5. Give one small exercise.
6. Review what they learned.
## Reader Mode Flow
1. Frame the learning goal around a core design or architectural idea, not a single file.
2. Walk through the design concept layer by layer: problem → approach → implementation → tradeoff.
3. Ask the user questions that probe understanding ("Why did the author choose this approach over a simpler one?"), not just prediction ("What happens next?").
4. Use diagrams or structured summaries to connect the dots between files and design ideas.
5. Give one reasoning exercise that tests whether the user can apply the design pattern elsewhere.
6. Review what they learned.
## Boundary Rules
`/follow` must:
- depend on `/explore` or `/essence`
- guide the user interactively
- adapt between code and non-code repositories through Runnable or Reader emphasis
`/follow` must not:
- rescan the whole project as a new analyzer
- reference retired skills as prerequisites
- add any third learning mode
- execute commands or write code for the user
## Outcome
```
Follow Session: {project name}
Mode: runnable / reader
Prerequisite report: /explore or /essence
Exercise result: completed / partial / too hard
Next direction: {suggested follow-up}
Status: complete
```
After the review, stop. Ask whether the user wants another exercise or wants to end the session.
-113
View File
@@ -1,113 +0,0 @@
# Environment Detection Rules
How to detect the runtime environment and guide the user through setup in `/follow`.
## Language Detection from Config
Check these files in order. The first match is the primary language.
| Config file | Language | Runtime check | Install command |
|---|---|---|---|
| `package.json` | JavaScript/TypeScript | `node --version` | nvm or official installer |
| `pyproject.toml` | Python | `python --version` | pyenv or python.org |
| `go.mod` | Go | `go version` | golang.org/dl |
| `Cargo.toml` | Rust | `rustc --version` | rustup |
| `pom.xml` | Java | `java -version` | SDKMAN or official |
| `build.gradle` / `build.gradle.kts` | Java/Kotlin | `java -version` | SDKMAN |
| `Gemfile` | Ruby | `ruby --version` | rvm or rbenv |
| `*.csproj` | C#/.NET | `dotnet --version` | .NET SDK |
| `CMakeLists.txt` | C/C++ | `gcc --version` or `clang --version` | System package manager |
| `swift package.json` | Swift | `swift --version` | Xcode or swift.org |
## Dependency Installation
Once language is detected, guide the user:
### JavaScript/TypeScript
```bash
# Check which package manager is used
if [ -f "yarn.lock" ]; then yarn install
elif [ -f "pnpm-lock.yaml" ]; then pnpm install
elif [ -f "bun.lockb" ] || [ -f "bun.lock" ]; then bun install
else npm install
fi
```
### Python
```bash
# Modern Python projects
pip install -e .
# Or with requirements
pip install -r requirements.txt
# Or with poetry
poetry install
# Or with uv
uv pip install -r requirements.txt
```
### Go
```bash
go mod download
```
### Rust
```bash
cargo build
```
### Java (Maven)
```bash
mvn install
```
### Java (Gradle)
```bash
./gradlew build
# or
gradle build
```
## Run Command Detection
How to start the project:
| Source | Command |
|---|---|
| `package.json` → `scripts.dev` | `npm run dev` |
| `package.json` → `scripts.start` | `npm start` |
| `Makefile` → `dev` target | `make dev` |
| `Makefile` → `run` target | `make run` |
| `pyproject.toml` (Poetry) | `poetry run python main.py` |
| `go.mod` → `package main` | `go run main.go` |
| `Cargo.toml` → `[[bin]]` | `cargo run` |
| `docker-compose.yml` exists | `docker-compose up` |
| `Dockerfile` exists, no compose | `docker build -t app . && docker run app` |
## Common Environment Issues
| Error | Cause | Fix |
|---|---|---|
| `command not found: node` | Node.js not installed | Install Node.js (recommend LTS) |
| `ModuleNotFoundError` | Python deps not installed | Run `pip install -r requirements.txt` |
| `EACCES: permission denied` | Global install without sudo | Use nvm/fnm, or prefix with sudo |
| `ENOENT: no such file` | Wrong working directory | `cd` to project root first |
| `port already in use` | Another process on same port | Kill the process or use different port |
| `go: cannot find main module` | Outside Go module | `cd` to directory with `go.mod` |
| `error: could not find Cargo.toml` | Outside Rust project | `cd` to directory with `Cargo.toml` |
| `java.lang.UnsupportedClassVersionError` | Wrong Java version | Match JDK version to project requirement |
| `npm ERR! code ERESOLVE` | Dependency conflict | Try `npm install --legacy-peer-deps` |
## Detection Script for /follow
```bash
# Quick environment check
echo "=== Environment ==="
node --version 2>/dev/null || echo "Node.js: not installed"
python --version 2>/dev/null || echo "Python: not installed"
go version 2>/dev/null || echo "Go: not installed"
rustc --version 2>/dev/null || echo "Rust: not installed"
java -version 2>/dev/null || echo "Java: not installed"
echo "PWD: $(pwd)"
```
Run this at the start of `/follow` Step 1 to understand what's available.