Review loop

.agents/skills/mpx2rn/references/rn-style-practice.md

@@ -478,13 +478,15 @@ px 和 rpx 在 RN 与小程序平台都具备良好兼容性，建议优先使
 
 **⚠️ 需要辅助属性的场景：**
 
-1. **`font-size` 的百分比**需要传递 `parent-font-size` 辅助属性
+1. **字体相关百分比**会优先基于当前或继承文本字号解析；没有可继承字号时按默认字号 `16` 计算
 2. **`calc()` 中出现的任何百分比**都需要传递相应的 `parent-width` / `parent-height` 辅助属性（`calc()` 是框架模拟支持的特性）
 
 ```html
 <template>
-  <!-- 场景1：font-size 百分比需要 parent-font-size -->
-  <view parent-font-size="{{16}}" class="text" />
+  <!-- 场景1：优先通过文本样式继承提供 font-size 百分比基准 -->
+  <view class="text-parent">
+    <text class="text">文本</text>
+  </view>
 
   <!-- 场景2：仅在无法替代时，calc() 中的百分比需要父级布局宽高辅助计算 -->
   <view id="calc-parent" class="calc-parent" wx:ref>
@@ -524,7 +526,11 @@ export default {
 
 <style>
   .text {
-    font-size: 120%; /* 需要 parent-font-size */
+    font-size: 120%; /* 优先基于继承字号；无继承字号时按默认字号 16 计算 */
+  }
+
+  .text-parent {
+    font-size: 16px;
   }
 
   .calc-parent {
@@ -549,9 +555,8 @@ export default {
 
 1. **优先使用 rpx**：对于固定尺寸，rpx 是最可靠的选择
 2. **放心使用百分比**：`width`, `height`, `padding`, `margin` 等属性的百分比由 RN 原生支持，可以放心使用
-3. **谨慎使用字体百分比**：`font-size` 的百分比需要 `parent-font-size` 辅助属性，建议使用 rpx 代替
-4. **谨慎使用 calc() 中的百分比**：该写法需要 `parent-width` / `parent-height` 辅助计算，通常还要查询父级布局并延迟展示，存在性能与体验开销；优先使用原生百分比、Flex、rpx / vw / vh 或固定尺寸替代
-5. **使用 vh/vw**：对于视口相关的尺寸，vh/vw 是更好的选择
+3. **谨慎使用 calc() 中的百分比**：该写法需要 `parent-width` / `parent-height` 辅助计算，通常还要查询父级布局并延迟展示，存在性能与体验开销；优先使用原生百分比、Flex、rpx / vw / vh 或固定尺寸替代
+4. **使用 vh/vw**：对于视口相关的尺寸，vh/vw 是更好的选择
 
 ### 1 像素边框（极细线）
 

.agents/skills/mpx2rn/references/rn-style-reference.md

@@ -205,10 +205,12 @@ mpx.config.rnConfig = Object.assign({}, mpx.config.rnConfig, {
 
 1.  **原生支持（基于父节点宽高）**： `width`、`height`、`left`、`right`、`top`、`bottom`、`margin`、`padding` 等属性。
 
-2.  **需辅助属性（基于父节点字体）**： `font-size` 的百分比计算依赖开发者传入的 `parent-font-size` 属性。
+2.  **文本字号相关百分比**：`font-size` / `line-height` 的百分比在文本样式透传阶段解析，基准优先使用当前已解析的 `font-size`、继承文本样式中的 `font-size`，若都不存在则使用默认字号 `16`。
 
     ```html
-    <text parent-font-size="16" style="font-size: 120%;">文本内容</text>
+    <view style="font-size: 16px;">
+      <text style="font-size: 120%;">文本内容</text>
+    </view>
     ```
 
 3.  **基于自身宽高计算**： `translateX`、`translateY`、`border-radius` 的百分比都是根据自身宽高来计算的（首次渲染不展示，在 `onLayout` 后计算生效）。其中 `translateX` / `translateY` 的百分比计算依赖 class 样式的编译转换；内联 `style` 中的字符串 transform 不会触发该计算。
@@ -559,7 +561,7 @@ Mpx 在 RN 平台支持 CSS 背景图及渐变背景，框架会自动处理样
 | `color` | `color` | - | 文本颜色 | `color: #333`；`color: rgba(0, 0, 0, 0.8)` |
 | `font` | `[font-style] [small-caps] [font-weight] font-size [ / line-height ] font-family` | - | 字体简写，展开为字体相关长属性；`font-size` 与 `font-family` 必填，`/` 两侧可有空格 | `font: italic bold 16px / 1.5 Arial` |
 | `font-family` | `string` | - | 字体（仅支持单字体），多字体 fallback 自动取首值并去除引号 | `font-family: PingFangSC-Regular` |
-| `font-size` | `length` | - | 字体大小 | `font-size: 28rpx`；`font-size: 16px` |
+| `font-size` | `length` \| `%` | - | 字体大小；百分比按文本透传字号基准解析 | `font-size: 28rpx`；`font-size: 16px`；`font-size: 120%` |
 | `font-weight` | `normal` \| `bold` \| `100-900` | `normal` | 字体粗细 | `font-weight: bold`；`font-weight: 500` |
 | `font-style` | `normal` \| `italic` | `normal` | 字体样式 | `font-style: italic` 斜体 |
 | `line-height` | `length` \| `number` \| `%` | - | 行高，纯数字值自动转换为百分比（如 `1.5` → `150%`） | `line-height: 40rpx`；`line-height: 1.5` |

.agents/skills/mpx2rn/references/rn-template-reference.md

@@ -640,7 +640,6 @@ Mpx 输出 RN 内置支持了大部分常用的基础组件，详情见下方文
 | enable-offset | boolean | `false` | 设置是否要获取组件的布局信息，若设置了该属性，会在 e.target 中返回组件的 offsetLeft、offsetWidth 信息 |
 | enable-var | boolean | `true` | 默认支持使用 css variable，若想关闭该功能可设置为 false |
 | enable-text-pass-through | boolean | `false` | RN 环境特有属性，开启文本样式和文本属性透传上下文，请在动态添加文本样式或文本属性时开启 |
-| parent-font-size | number |  | 父组件字体大小，主要用于百分比计算的场景，如 font-size: 100% |
 | parent-width | number |  | 父组件宽度，主要用于百分比计算的场景，如 width: calc(100% - 20px)，需要在外部传递父组件的宽度 |
 | parent-height | number |  | 父组件高度，主要用于百分比计算的场景，如 height: calc(100% - 20px),需要在外部传递父组件的高度 |
 

.agents/skills/review-loop/SKILL.md

@@ -0,0 +1,70 @@
+---
+name: review-loop
+description: Use this skill whenever the user asks for a review loop, self-review workflow, planner/reviewer/coder/code-reviewer collaboration, plan-before-code process, multi-round agent review, or a stable subagent workflow that must produce a plan, wait for user confirmation, implement code, review the diff, and preserve revision records. This skill requires real subagent support and must not fall back to single-agent roleplay.
+protocolVersion: 1.0.0
+---
+
+# Review Loop
+
+Use this skill to run a two-phase workflow:
+
+1. Plan loop: `planner` writes or revises `plan.md`; `plan-reviewer` reviews it.
+2. Code loop: after user confirmation, `coder` implements the confirmed plan; `code-reviewer` reviews the resulting diff.
+
+The workflow requires real subagents. If the current environment cannot create real subagents, stop and tell the user that `review-loop` cannot run in this session.
+
+## Required Setup
+
+Before doing task work:
+
+1. Read `references/protocol.md`, `references/state-machine.md`, and `references/role-contracts.md`.
+2. Read `references/codex.md` or `references/claude-code.md` when that platform applies.
+3. Create a task id.
+4. Run `scripts/init-workspace.js` to create `.agent-workflows/review-loop/<task-id>/`.
+5. Run `scripts/prepare-agent-roles.js` to prepare real subagent roles.
+6. Run `scripts/validate-state.js` before advancing.
+
+Default `maxRounds` is `3`. If the user specifies a maximum loop count, pass it to `init-workspace.js`.
+
+## Hard Rules
+
+- Do not implement code until the plan loop has ended and the user explicitly confirms the plan.
+- Do not let reviewer roles modify `plan.md` or source files; reviewers write JSON review files only.
+- Use `scripts/validate-review-json.js` after every reviewer output.
+- Use `scripts/snapshot-diff.js` after every code loop implementation round.
+- Use `scripts/advance-state.js` for every state transition; never hand-edit `state.json`.
+- Stop a loop when the reviewer returns `approved` or when the loop reaches `state.json.maxRounds`.
+- When a loop stops because `maxRounds` is reached, summarize remaining findings and ask the user whether to continue, accept, or revise.
+- Preserve every round's revision record in `plan.md`.
+
+## Subagent Roles
+
+The four required roles are:
+
+- `planner`
+- `plan-reviewer`
+- `coder`
+- `code-reviewer`
+
+Use the templates under `templates/roles/` to prepare role definitions. If project-level roles already exist, reuse them. If not, ask whether to create temporary roles under the task workspace or persist roles to the project agent config.
+
+## User Confirmation Gates
+
+After the plan loop ends, show the user:
+
+- `plan.md` path
+- plan loop termination reason
+- review summary by round
+- unaccepted findings and reasons
+
+Ask whether to enter the code loop. Only after explicit confirmation, run `advance-state.js --event confirm-plan`.
+
+After the code loop ends, show the user:
+
+- `plan.md` path
+- diff summary
+- code loop termination reason
+- validation commands and results
+- remaining risks
+
+Ask whether to accept the final result. Only after explicit confirmation, run `advance-state.js --event confirm-final`.

.agents/skills/review-loop/evals/evals.json

@@ -0,0 +1,23 @@
+{
+  "skill_name": "review-loop",
+  "evals": [
+    {
+      "id": 1,
+      "prompt": "使用 review-loop 工作流实现一个小的 bugfix，先产出方案并 review，等我确认后再写代码。",
+      "expected_output": "Creates a review-loop workspace, initializes state.json with maxRounds=3, enters plan loop, and waits for user confirmation before code loop.",
+      "files": []
+    },
+    {
+      "id": 2,
+      "prompt": "使用 review-loop 工作流实现需求，最多 review 5 轮。",
+      "expected_output": "Initializes state.json with maxRounds=5 and uses that value to stop plan/code loops.",
+      "files": []
+    },
+    {
+      "id": 3,
+      "prompt": "继续上次中断的 review-loop 任务。",
+      "expected_output": "Runs recoverability/state checks, reads state.json, and resumes from the recorded phase instead of starting over.",
+      "files": []
+    }
+  ]
+}

.agents/skills/review-loop/references/claude-code.md

@@ -0,0 +1,51 @@
+# Claude Code Integration
+
+Claude Code support requires real subagents. Do not run this workflow as single-agent roleplay.
+
+## Role Discovery
+
+Check project agents first:
+
+```text
+.claude/agents/planner.md
+.claude/agents/plan-reviewer.md
+.claude/agents/coder.md
+.claude/agents/code-reviewer.md
+```
+
+If all four exist, reuse them.
+
+If any are missing, ask the user to choose:
+
+- temporary roles under `.agent-workflows/review-loop/<task-id>/runtime/roles/`
+- persistent project agents under `.claude/agents/`
+
+Use `scripts/prepare-agent-roles.js` for either path.
+
+## Temporary Roles
+
+Temporary roles are copied from `templates/roles/` to:
+
+```text
+.agent-workflows/review-loop/<task-id>/runtime/roles/
+```
+
+They are valid only for the current task.
+
+## Persistent Roles
+
+Persistent roles are copied from `templates/roles/` to:
+
+```text
+.claude/agents/
+```
+
+This requires explicit user confirmation because it modifies project-level agent configuration.
+
+## Failure
+
+If Claude Code cannot create real subagents in the current session, stop the workflow and tell the user:
+
+```text
+review-loop requires real subagent support. Current Claude Code session cannot create subagents, so the workflow cannot continue.
+```

.agents/skills/review-loop/references/codex.md

@@ -0,0 +1,57 @@
+# Codex Integration
+
+Codex support requires real subagents. Do not run this workflow as single-agent roleplay.
+
+## Role Discovery
+
+Check project custom agents first:
+
+```text
+.codex/agents/planner.toml
+.codex/agents/plan-reviewer.toml
+.codex/agents/coder.toml
+.codex/agents/code-reviewer.toml
+```
+
+If all four exist, reuse them.
+
+If any are missing, ask the user to choose:
+
+- temporary roles under `.agent-workflows/review-loop/<task-id>/runtime/roles/`
+- persistent project custom agents under `.codex/agents/`
+
+Use `scripts/prepare-agent-roles.js` for either path.
+
+## Persistent Agent Format
+
+Codex custom agents are TOML files. Each file must define:
+
+```toml
+name = "planner"
+description = "Planner for review-loop workflows."
+developer_instructions = """
+Read the review-loop workspace inputs and produce or revise plan.md.
+Do not implement code.
+"""
+```
+
+## Starting Roles
+
+Codex does not spawn subagents without an explicit request. After roles are prepared, the orchestrator must explicitly start the four role subagents as the state machine requires.
+
+For each role, include:
+
+- role name
+- task id
+- workspace path
+- input files
+- required output path
+- instruction to return a concise summary to the orchestrator
+
+## Failure
+
+If Codex cannot create real subagents in the current session, stop the workflow and tell the user:
+
+```text
+review-loop requires real subagent support. Current Codex session cannot create subagents, so the workflow cannot continue.
+```