P3b: remove workflow-learning/replay from active prompt perimeter

- Codex/Skills CN+EN AGENTS.md: remove workflow-learning strategy block,
  routing row, skill table row, replay/ dir entry, footer replay ref
- Claude/Skills CN+EN CLAUDE.md: same symmetric cleanup
- skills/sopify/workflow-learning/ (×4): delete from install shipped surface
- tests/test_installer.py: align footer_contract_line assertions (L1302, L1356)
- .sopify-skills/project.md: align test dialect (L27 unittest→pytest,
  L30 bundle smoke unittest→pytest), drop replay/handoff paired term (L34)
- tasks.md 2.13: mark done

Release-Sync: yes

Release-Version: 2026-05-08.191000

Release-Date: 2026-05-08

Author: Li-Sanze

.sopify-skills/history/2026-05/20260508_p3b_perimeter_cleanup/tasks.md

@@ -44,7 +44,7 @@ plan_status: completed
 ### 旧概念清理（P3a 已 sunset surface 残留）
 
 - [x] 2.12 tests 中验证 P3a 已 sunset surface 的断言——更新或删除
-- [ ] 2.13 prompt 中引用已 sunset contract 的段落——清除（prompt asset 本体含 replay 字面量，待 prompt asset 整体更新时处理）
+- [x] 2.13 prompt 中引用已 sunset contract 的段落——清除（Codex/Skills + Claude/Skills CN|EN：AGENTS/CLAUDE.md 路由表/策略块/技能行、footer replay 引用、replay/ 目录条目均已移除；skills/sopify/workflow-learning/ 目录（安装链路 shipped surface）全部删除）
 - [x] 2.14 handoff/output 旧兼容投影——清除
 - [x] 2.15 reason phrasing / phase label 特判——清除
 

.sopify-skills/project.md

@@ -24,14 +24,14 @@
 - `runtime/models.py` 是稳定公开 facade；`from runtime.models import X` 继续作为对外兼容入口。
 - 具体实现收敛到 `runtime/_models/`，当前按 `core / decision / artifacts / summary / handoff` 分组，避免在公开路径下继续堆积单文件复杂度。
 - facade 必须维护显式 `__all__`，保证 `from runtime.models import *` 的 surface 仍然可控。
-- repo-local runtime 回归统一使用 `python3 -m unittest discover tests -v`，避免拆分后因手写文件列表漏测。
+- repo-local runtime 回归统一使用 `python3 -m pytest tests -v`，避免拆分后因手写文件列表漏测。
 - repo-local 共享测试 helper 固定收敛到 `tests/runtime_test_support.py`；`tests/test_runtime_*.py` 负责按主题拆分具体 `TestCase`。
 - bundle 对外继续保留 `.sopify-runtime/tests/test_runtime.py` 路径，但该文件只承担最小 smoke contract，不再复制 repo-local 全量 runtime 测试。
-- 需要对绝对路径下的 bundle smoke 做便携校验时，统一使用 `python3 -m unittest discover -s <bundle-tests-dir> -p 'test_runtime.py' -v`，避免 `unittest` 把绝对路径误当成模块名。
+- 需要对绝对路径下的 bundle smoke 做便携校验时，统一使用 `python3 -m pytest <bundle-tests-dir>/test_runtime.py -v`，避免路径解析歧义。
 
 ## Develop 质量约定
 
-- `continue_host_develop` 仍是宿主负责真实代码修改的正式模式；runtime 只负责 machine-readable quality contract、checkpoint callback 与 replay/handoff 落盘。
+- `continue_host_develop` 仍是宿主负责真实代码修改的正式模式；runtime 只负责 machine-readable quality contract、checkpoint callback 与 handoff 落盘。
 - develop 质量循环的正式发现顺序固定为：`.sopify-skills/project.md verify` > 项目原生脚本/配置 > `not_configured` 可见降级。
 - develop 质量结果的正式字段固定为：`verification_source / command / scope / result / reason_code / retry_count / root_cause / review_result`。
 - `result` 的稳定值域固定为：`passed / retried / failed / skipped / replan_required`；`root_cause` 的稳定值域固定为：`logic_regression / environment_or_dependency / missing_test_infra / scope_or_design_mismatch`。

CHANGELOG.md

@@ -6,6 +6,21 @@ Format: Summary → Changed → Plan Packages. File-level details live in `git l
 
 ## [Unreleased]
 
+## [2026-05-08.191000] - 2026-05-08
+
+### Summary
+
+- Archived 1 plan package(s); Changes across: Skills, Tests.
+
+### Changed
+
+- **Skills**: Synced prompt-layer skills (12 files)
+- **Tests**: Updated automated coverage (1 files)
+
+### Plan Packages
+
+- `20260508_p3b_perimeter_cleanup` (archived)
+
 ## [2026-05-07.220011] - 2026-05-07
 
 ### Summary

Claude/Skills/CN/CLAUDE.md

@@ -1,5 +1,5 @@
 <!-- bootstrap: lang=zh-CN; encoding=UTF-8 -->
-<!-- SOPIFY_VERSION: 2026-05-07.220011 -->
+<!-- SOPIFY_VERSION: 2026-05-08.191000 -->
 <!-- ARCHITECTURE: Adaptive Workflow + Layered Rules -->
 
 # Sopify - 自适应 AI 编程助手
@@ -69,7 +69,7 @@ Next: {下一步提示}
 **Footer 契约：**
 - footer 固定跟在 `Changes` 区块之后
 - `Next:` 必须作为 footer 最后一行。
-- footer 不展示生成时间；若需要机器可审计时间戳，内部摘要 / replay 文件可继续使用 ISO 8601（可带时区）。
+- footer 不展示生成时间；若需要机器可审计时间戳，内部摘要文件可继续使用 ISO 8601（可带时区）。
 
 **状态符：**
 | 符号 | 含义 |
@@ -142,20 +142,6 @@ Next: {下一步提示}
 说明：当 `current_handoff.json.required_host_action == continue_host_develop` 时，宿主继续负责真实代码修改；但若开发中再次出现“需要用户补事实 / 拍板选路”的分叉，宿主不得自由追问，也不得手写 `current_decision.json / current_handoff.json`，而必须调用 `scripts/develop_callback_runtime.py submit --payload-json ...`（vendored 对应 `.sopify-runtime/scripts/develop_callback_runtime.py`）回调 runtime。payload 必须包含 `checkpoint_kind` 与 `resume_context`；当前 `resume_context` 至少要求 `active_run_stage / current_plan_path / task_refs / changed_files / working_summary / verification_todo`。
 说明：当 `current_handoff.json.required_host_action == continue_host_consult` 时，宿主只可在已消费当前回合 gate contract 的前提下继续问答；不得在 gate 前自行路由，也不得在 gate 后再次重判 consult / 非 consult。宿主的回答应基于当前 gate contract 与 `handoff.artifacts` 暴露的 consult context（如有）生成；若缺少额外 consult context，应显式按当前请求降级回答，而不是用宿主侧语义分析补出另一条路由。
 
-**workflow-learning 主动记录策略：**
-```yaml
-workflow:
-  learning:
-    auto_capture: by_requirement # always | by_requirement | manual | off
-```
-
-| 值 | 行为 |
-|-----|------|
-| `always` | 所有开发任务主动记录（full） |
-| `by_requirement` | 按复杂度主动记录：simple=off，medium=summary，complex=full |
-| `manual` | 仅在用户明确要求“开始记录这次任务”后记录 |
-| `off` | 不主动新建记录；但回放/复盘意图识别与已有记录回放仍可用 |
-
 ---
 
 ## Auto Rules (自动规则)
@@ -220,7 +206,6 @@ workflow:
 │   ├── preferences.md
 │   └── feedback.jsonl
 ├── project.md               # 技术约定，不与 background/design 重复
-└── replay/                  # 可选回放能力，继续忽略
 ```
 
 ### A6 | 生命周期管理
@@ -290,7 +275,6 @@ progressive: 按需创建文件 (默认)
 | 路由 | 条件 | 行为 |
 |-----|------|-----|
 | 咨询问答 | 纯问题，无代码变更 | 先过 gate，再按 consult handoff 由宿主回答 |
-| 复盘学习 | 提到回放/复盘/为什么这么做（意图识别始终开启） | 调用 workflow-learning，生成记录与讲解 |
 | 快速修复 | ≤2 文件，明确修改 | 直接执行 |
 | 轻量迭代 | 3-5 文件，清晰需求 | light 方案 + 执行 |
 | 完整开发 | >5 文件或架构变更 | 3 阶段完整流程 |
@@ -432,7 +416,6 @@ Next: 请验证功能
 | `develop` | 进入开发实施 | 代码执行、KB同步 |
 | `kb` | 知识库操作 | 初始化、更新策略 |
 | `templates` | 创建文档 | 所有模板定义 |
-| `workflow-learning` | 用户要求回放/复盘/原因讲解，或 `auto_capture` 命中主动记录策略 | 完整记录、回放、逐步讲解 |
 
 **读取方式：** 按需读取，进入对应阶段时加载。
 

Claude/Skills/CN/skills/sopify/workflow-learning/CHANGELOG.md

@@ -1,5 +1,5 @@
 <!-- bootstrap: lang=en-US; encoding=UTF-8 -->
-<!-- SOPIFY_VERSION: 2026-05-07.220011 -->
+<!-- SOPIFY_VERSION: 2026-05-08.191000 -->
 <!-- ARCHITECTURE: Adaptive Workflow + Layered Rules -->
 
 # Sopify - Adaptive AI Programming Assistant
@@ -69,7 +69,7 @@ Next: {Next step hint}
 **Footer Contract:**
 - the footer always follows the `Changes` block
 - `Next:` must be the final line in the footer.
-- the footer does not display generated time; if a machine-auditable timestamp is needed, internal summary / replay artifacts may keep ISO 8601 timestamps with timezone data.
+- the footer does not display generated time; if a machine-auditable timestamp is needed, internal summary files may keep ISO 8601 timestamps with timezone data.
 
 **Status Symbols:**
 | Symbol | Meaning |
@@ -142,20 +142,6 @@ Note: when `current_handoff.json.required_host_action == confirm_decision`, the
 Note: when `current_handoff.json.required_host_action == continue_host_develop`, the host still owns real code changes; but if implementation hits another user-facing branch, the host must not ask a free-form question or hand-write `current_decision.json / current_handoff.json`. It must call `scripts/develop_callback_runtime.py submit --payload-json ...` instead (vendored: `.sopify-runtime/scripts/develop_callback_runtime.py`). The payload must contain `checkpoint_kind` plus `resume_context`; the current minimum `resume_context` fields are `active_run_stage / current_plan_path / task_refs / changed_files / working_summary / verification_todo`.
 Note: when `current_handoff.json.required_host_action == continue_host_consult`, the host may continue the discussion only after consuming the current turn's gate contract. It must not self-route before the gate, and it must not re-decide consult vs non-consult after the gate. The answer must be grounded in the current gate contract plus any consult context exposed by `handoff.artifacts`; if that context is missing, degrade explicitly against the current request instead of inferring a different route from chat semantics.
 
-**workflow-learning proactive capture policy:**
-```yaml
-workflow:
-  learning:
-    auto_capture: by_requirement # always | by_requirement | manual | off
-```
-
-| Value | Behavior |
-|------|----------|
-| `always` | Proactively capture all development tasks (full) |
-| `by_requirement` | Capture by complexity: simple=off, medium=summary, complex=full |
-| `manual` | Capture only after explicit request like "start recording this task" |
-| `off` | Do not proactively create new logs; replay/review intent and replay from existing sessions still work |
-
 ---
 
 ## Auto Rules
@@ -220,7 +206,6 @@ Complex: Files > 5, architectural changes, new features
 │   ├── preferences.md
 │   └── feedback.jsonl
 ├── project.md               # Technical conventions, not a duplicate of background/design
-└── replay/                  # Optional replay capability, still ignored
 ```
 
 ### A6 | Lifecycle Management
@@ -290,7 +275,6 @@ Semantic analysis routing:
 | Route | Condition | Behavior |
 |-------|-----------|----------|
 | Q&A | Pure question, no code changes | Run gate first, then answer through consult handoff in the host session |
-| Workflow Learning | Mentions replay/review/why this choice (intent recognition is always enabled) | Call workflow-learning for trace capture and explanation |
 | Quick Fix | ≤2 files, clear modification | Direct execution |
 | Light Iteration | 3-5 files, clear requirements | Light plan + execution |
 | Full Development | >5 files or architectural changes | Full 3-phase workflow |
@@ -432,7 +416,6 @@ Next: Please verify the functionality
 | `develop` | Enter development | Code execution, KB sync |
 | `kb` | Knowledge base operations | Init, update strategies |
 | `templates` | Create documents | All template definitions |
-| `workflow-learning` | User asks replay/review/why, or `auto_capture` proactively applies | Full trace logging, replay, step-by-step explanation |
 
 **Loading:** On-demand, loaded when entering corresponding phase.