@hupan56/wlkj 2.0.0

package/templates/qoder/commands/wl-prd.md

@@ -0,0 +1,199 @@
+---
+name: wl-prd
+description: "Generate PRD with prototype matching project style"
+argument-hint: "[mode] [description]  modes: reference/brainstorm/plan/quick"
+auto-approve: true
+allowed-tools: [Read, Glob, Grep, Bash, Write, Edit, Agent]
+---
+
+# /wl-prd - Generate PRD + Prototype
+
+User input: $ARGUMENTS
+
+## ⚠️ STEP 0: MANDATORY PLATFORM SELECTION (MUST DO FIRST!)
+
+**Before ANY search, ANY file read, ANY analysis — ask this question.**
+**NEVER auto-detect. NEVER assume. NEVER skip. ALWAYS ASK.**
+
+```
+这个需求是针对哪个平台？
+
+1. **Web 管理端** (fywl-ui) - Ant Design Vue + VxeGrid 风格
+2. **APP 移动端** (Carmg-H5) - Vant 风格
+3. **两端都要**
+
+请选择 (1/2/3):
+```
+
+WAIT for the answer. Platform determines everything downstream:
+
+| Answer | Platform | Project | Search Flag | Prototype Template |
+|--------|----------|---------|-------------|--------------------|
+| 1 / Web / PC / 管理端 | Web | fywl-ui | `--platform web` | .qoder/templates/prototype-web.html |
+| 2 / APP / H5 / 移动端 | APP | Carmg-H5 | `--platform app` | .qoder/templates/prototype-app.html |
+| 3 / 都要 / 两端 | Both | Both | run both | both templates, 2 prototypes |
+
+## Mode Detection
+
+| Keyword in input | Mode | Sub-agent (.qoder/agents/) |
+|------------------|------|---------------------------|
+| (none, DEFAULT) | Reference | prd-reference.md |
+| "头脑风暴" / "brainstorm" | Brainstorm | prd-research.md (N parallel) |
+| "规划" / "plan" | Planning | prd-planning.md |
+| "快速" / "quick" | Quick | prd-quick.md |
+
+The agent files define the detailed workflow for each mode. When
+dispatching a sub-agent, ALWAYS include in the dispatch prompt:
+the chosen platform, the user requirement, and the current developer name.
+For simple requirements you may run the mode workflow inline instead of
+dispatching — the steps are the same.
+
+## Storage
+
+Draft -> workspace/members/{developer}/drafts/REQ-{ID}-{desc}.md
+Confirmed -> workspace/specs/prd/REQ-{ID}-{desc}.md
+Prototype (single platform) -> workspace/members/{developer}/drafts/prototype-{feature}.html
+Prototype (Both) -> prototype-{feature}-web.html AND prototype-{feature}-app.html
+
+REQ-ID rule: **必须**调用原子分配器获取新 ID, 禁止手动扫描最大值 (多人并发会撞号):
+```bash
+python -c "import sys; sys.path.insert(0,'.qoder/scripts'); from common.reqid import allocate_req_id; n=allocate_req_id(); print('REQ-%d-%03d' % (__import__('datetime').date.today().year, n))"
+```
+分配器在文件锁内读 data/index/req-counter.json + 递增 + 写回, 保证并发安全。
+首次使用若计数器不存在, 会自动 migrate_from_existing() 扫现有 PRD 初始化。
+Never reuse an existing number. push 时 team_sync 会校验 REQ-ID 唯一性。
+
+---
+
+## Reference Mode (DEFAULT)
+
+### Step 1: Get ALL Context in ONE Call (fast path)
+```bash
+python .qoder/scripts/context_pack.py <业务关键词> --platform <web|app> [--type table|form|dashboard]
+```
+One command returns: related code files + same-type page examples +
+real field specs + related PRDs + APIs + style/icon/template pointers.
+Then read the top 2-3 most similar Vue files it lists.
+(Fallback: individual search_index.py calls if you need to drill deeper.)
+
+### Step 2: Read History PRD
+Read related PRDs found by `--prd` from data/docs/prd/.
+
+### Step 3: Read Business Drafts
+Read data/docs/drafts/ for business context.
+
+### Step 4: Style Tokens
+- Web: read data/index/vben-style-reference.json (pre-distilled design tokens)
+- APP: Vant variables (see Prototype Rules below)
+- Only open data/style/ PDFs if a needed token is missing from the JSON.
+- If PDF conflicts with real code, CODE WINS.
+
+### Step 5: Confirm Findings (BATCHED)
+Present ALL findings in ONE numbered list (existing behavior, related pages,
+fields, business rules, assumptions). Ask the user to confirm or correct
+in a single round. Only ask follow-up questions for genuinely ambiguous points.
+Do NOT ask one question per finding.
+
+### Step 6: Generate PRD + Prototype
+
+**PRD**: Auto-detect product type, use template, run 10-point quality check
+(see prd-generator skill).
+
+**Prototype** (KEY FEATURE):
+1. Determine prototype type (table/form/detail/modal)
+2. Start from the platform template (.qoder/templates/prototype-web.html or prototype-app.html)
+3. Fill placeholders with real fields/columns extracted from existing code
+4. Changed/new points highlighted (highlight-new class)
+5. Clickable, fillable, hover effects; single self-contained file
+6. Save per Storage rules above (Both -> 2 files with -web/-app suffix)
+
+### Step 7: EVA Gate + Confirm Publish + Auto Sync
+Before asking to publish, run the quality gate:
+```bash
+python .qoder/scripts/eval_prd.py <draft-prd.md> <prototype.html>
+```
+If it FAILs (<80%), fix the reported issues (missing sections / invented
+fields / non-canonical colors) and re-run until PASS. Then ask user to
+confirm publishing to workspace/specs/prd/.
+After publishing (and also after saving drafts/prototypes), run:
+```bash
+python .qoder/scripts/team_sync.py push
+```
+This silently commits & pushes the PM's outputs to the team repo.
+The user NEVER needs to know git. If the script prints SYNC_CONFLICT,
+YOU (the AI) resolve the git conflict following its instructions —
+do not ask the user to run git commands.
+
+---
+
+## Brainstorm Mode
+
+1. SCAMPER 7 dimensions + 5-Why root cause analysis (inline)
+2. Dispatch N sub-agents (default 2-4, scale with topic size) using
+   .qoder/agents/prd-research.md — each gets a distinct perspective
+   (competitor / user / tech / risk), web research only
+3. Each agent writes workspace/members/{developer}/drafts/brainstorm-agent{N}-{perspective}.md
+4. MERGE all agent reports into one summary, present to user,
+   then DELETE the per-agent brainstorm-agent*.md files
+5. Generate PRD + prototype matching target platform style (Step 6 above)
+
+---
+
+## Planning Mode
+
+Follow .qoder/agents/prd-planning.md:
+1. Current state analysis (read code, platform-aware search)
+2. Ideal state research (web search)
+3. Gap analysis table
+4. Current plan (MVP) + Future plan
+5. ONE confirmation round on the full report (adjust at most twice, don't loop forever)
+6. Output plan-{topic}.md AND (if user wants to proceed) PRD + prototype via Step 6
+
+---
+
+## Quick Mode
+
+Follow .qoder/agents/prd-quick.md:
+1. Quick locate code (platform-aware search)
+2. Generate Mini-PRD + quick prototype (only change the diff)
+3. One confirmation
+
+---
+
+## Data Reading Matrix
+
+| Mode | Read Code | Read PRD | Web Search | Read Style | Generate Prototype |
+|------|-----------|----------|------------|------------|-------------------|
+| Reference | YES | YES | YES | YES | YES match |
+| Brainstorm | STYLE ONLY* | NO | YES | YES | YES new |
+| Planning | YES | YES | YES | YES | YES diff |
+| Quick | YES | NO | NO | YES | YES micro |
+
+*Brainstorm: business research happens on the web, but the prototype style
+MUST come from real code. Before generating any new-feature prototype:
+run `--style <type> --platform <p>`, read 1-2 closest real pages, and for
+dashboard/看板/大屏 requirements read data/index/chart-style-reference.json
+(two in-system idioms: 大屏深蓝风 = dashboard/monitor, 浅色卡片风 =
+dashboard/workbench). Never invent a generic admin look.
+
+---
+
+## Prototype Rules
+
+1. Style priority: real .vue code > vben-style-reference.json / Vant tokens > PDF spec
+2. Only change the diff, everything else identical to existing page
+3. Interactive: buttons, forms, tables must work
+4. Single file: one HTML with all CSS and JS
+5. Templates already contain the correct design tokens — do not invent colors
+
+### Web (fywl-ui) tokens — from data/index/vben-style-reference.json
+- Primary: `hsl(212 100% 45%)` via `--primary` CSS variable
+- Destructive: `hsl(359.33 100% 65.1%)`, Success: `hsl(144 57% 58%)`
+- Background: `hsl(216 20.11% 95.47%)` deep / #ffffff card
+- Font: PingFang SC / Microsoft YaHei, 14px
+- Table: VxeGrid style, fixed header, stripe
+- Form: labelWidth 80-120px, grid 2-4 columns
+
+### APP (Carmg-H5) tokens — Vant 3
+- `--van-primary-color: #1989fa`, success #07c160, warning #ff976a, danger #ee0a24
+- max-width: 375px, bottom tab bar, mobile layout

package/templates/qoder/commands/wl-report.md

@@ -0,0 +1,166 @@
+---
+name: wl-report
+description: "Generate daily or weekly work report. Auto-isolates by current developer."
+argument-hint: "[daily/weekly] or empty for auto-detect"
+auto-approve: true
+allowed-tools: [Read, Bash]
+---
+
+# /wl-report - Work Report (Isolated by Developer)
+
+User input: $ARGUMENTS
+
+## CRITICAL: Data Isolation
+
+This report ONLY shows data for the CURRENT developer.
+Read `.qoder/.developer` to get the current developer name.
+ALL data sources are filtered by this name.
+
+If no developer is set, refuse and say "Run /wl-init first".
+
+## Mode Detection
+
+- "daily" or "日报" -> Daily report
+- "weekly" or "周报" -> Weekly report
+- No keyword + today is Friday -> Weekly report
+- No keyword + other days -> Daily report
+
+## Data Sources (All Filtered by Current Developer)
+
+### 1. Who am I?
+```
+Read .qoder/.developer -> get developer name (e.g., "小王")
+This name is used to filter ALL data below.
+```
+
+### 2. My Tasks (workspace/tasks/)
+```bash
+# Find tasks where creator OR assignee = current developer
+python .qoder/scripts/task.py list --mine
+```
+Only show tasks where task.json has:
+- `creator` == current developer, OR
+- `assignee` == current developer
+
+### 3. My Session History
+```
+Read: workspace/members/{current_developer}/journal/
+Only files in THIS developer's journal directory.
+```
+
+### 4. My Git Commits
+```bash
+# For each project in data/code/:
+cd data/code/{project}
+git log --author="{current_developer}" --since="{date}" --oneline
+```
+Filter by git author matching current developer name.
+
+### 5. My PRDs
+```
+Read: workspace/members/{current_developer}/drafts/ -> drafts by this developer
+Read: workspace/specs/prd/*.md -> check if PRD frontmatter has author == current developer
+```
+
+---
+
+## Daily Report Flow
+
+1. Read `.qoder/.developer` -> get developer name and role
+2. Calculate today's date range
+3. Filter tasks by creator/assignee == developer
+4. Filter git log by --author == developer
+5. Read developer's journal entries from today
+6. Check developer's drafts for new PRDs
+7. Generate report with ONLY this developer's data
+
+Output:
+```
+# 日报 - {developer}({role}) - {date}
+
+## 今日完成
+- [Task] 完成了 XXX
+- [PRD] 撰写了 XXX
+- [Code] 提交了 XXX
+
+## 进行中
+- [Task] XXX (进度 XX%)
+
+## 明日计划
+- 基于当前任务状态自动建议
+
+## 备注
+- {any issues or blockers}
+```
+
+Save to: workspace/members/{developer}/journal/daily-{date}.md
+Then auto-sync (user never touches git): `python .qoder/scripts/team_sync.py push`
+
+---
+
+## Weekly Report Flow
+
+1. Read `.qoder/.developer` -> get developer
+2. Calculate this week's range (Monday to today)
+3. Aggregate:
+   - Tasks completed this week (creator/assignee == developer)
+   - Git commits this week (--author == developer)
+   - PRDs created this week
+   - Session count from journal
+4. Calculate stats:
+   - Tasks completed / total tasks
+   - Lines of code (from git diff --stat)
+   - PRDs written
+5. Generate weekly summary
+
+Output:
+```
+# 周报 - {developer}({role}) - {week_range}
+
+## 本周概览
+| 指标 | 数量 |
+|------|------|
+| 完成任务 | N |
+| 进行中任务 | N |
+| Git提交 | N次 |
+| PRD产出 | N份 |
+
+## 本周完成 (按天分组)
+### 周一
+- ...
+
+### 周二
+- ...
+
+## 未完成/延期
+- {overdue items, only this developer's}
+
+## 下周计划
+- {based on pending tasks assigned to this developer}
+
+## 问题与建议
+- {any patterns or issues}
+```
+
+Save to: workspace/members/{developer}/journal/weekly-{week}.md
+Then auto-sync (user never touches git): `python .qoder/scripts/team_sync.py push`
+
+---
+
+## Why This Is Isolated
+
+| Data Source | Filter Method |
+|-------------|--------------|
+| Tasks | task.json `creator` or `assignee` == developer |
+| Journal | Only read from `workspace/members/{developer}/journal/` |
+| Git | `git log --author="{developer}"` |
+| PRD Drafts | Only read from `workspace/members/{developer}/drafts/` |
+| Confirmed PRDs | Check frontmatter `author` field |
+
+Each developer has their OWN:
+- drafts/ folder
+- inbox/ folder  
+- journal/ folder
+- task assignments
+
+No data from other developers is included.

package/templates/qoder/commands/wl-search.md

@@ -0,0 +1,52 @@
+---
+name: wl-search
+description: "Search code index by keyword or API. Auto-approve. Returns compact results."
+argument-hint: "[keyword] or --api [path] or --list or --modules"
+auto-approve: true
+allowed-tools: [Read, Bash]
+---
+
+# /wl-search - Search Code Index
+
+User input: $ARGUMENTS
+
+## THIS IS THE ONLY WAY TO SEARCH CODE
+
+When user asks about ANY code/business/API, use this command.
+It calls search_index.py which returns compact results (~200 bytes).
+DO NOT grep the entire codebase.
+
+## Execution
+
+Run exactly this command:
+```
+python .qoder/scripts/search_index.py $ARGUMENTS
+```
+
+If no arguments provided, run:
+```
+python .qoder/scripts/search_index.py --list
+```
+
+## How to Use Results
+
+The script returns:
+- Matched files grouped by project
+- Shortened paths (project/module/.../file.java)
+- Max 5 files per keyword
+
+After getting results:
+1. Pick the most relevant files (2-3 max)
+2. Read ONLY those files directly
+3. Answer the user's question
+
+## Examples
+
+- `/wl-search 考勤` -> searches attendance related files
+- `/wl-search insurance --platform web` -> insurance files in fywl-ui only
+- `/wl-search --prd 保险` -> searches collected PRDs
+- `/wl-search --style table --platform web` -> table-page examples in fywl-ui
+- `/wl-search --field nickName` -> where a field is used (label + files)
+- `/wl-search --api salary` -> searches salary API endpoints
+- `/wl-search --modules` -> shows project overview
+- `/wl-search --list` -> shows top 50 keywords

package/templates/qoder/commands/wl-spec.md

@@ -0,0 +1,18 @@
+---
+name: wl-spec
+description: "Generate or review Spec from PRD + Design"
+argument-hint: "[review]"
+auto-approve: true
+allowed-tools: [Read, Glob, Grep, Bash, Write, Edit]
+---
+
+# /wl-spec - Generate Spec
+
+User input: 
+
+## Auto Generate Spec
+1. Read current task PRD and design artifacts
+2. Load .qoder/agents/spec-generator.md
+3. Generate spec.md based on PRD + Design
+4. Save to workspace/specs/ or workspace/tasks/{id}/spec.md
+5. Notify dev to review

package/templates/qoder/commands/wl-status.md

@@ -0,0 +1,51 @@
+---
+name: wl-status
+description: "Project status with roadmap, milestones, dependency tracking, and health score"
+argument-hint: "[scope: current/roadmap/health]"
+auto-approve: true
+allowed-tools: [Read, Glob, Grep, Bash]
+---
+
+# /wl-status - Project Status + Roadmap
+
+User input: $ARGUMENTS (default: show current status)
+
+## Scopes
+- (no arg or "current") -> Current sprint status
+- "roadmap" -> Now/Next/Later roadmap view
+- "health" -> Health score assessment
+
+## Current Status (DEFAULT)
+Show:
+1. Active developer (from .qoder/.developer)
+2. Current task (if any) + state
+3. Sprint overview:
+   | Metric | Value |
+   | Total tasks | N |
+   | In progress | N |
+   | Completed | N |
+   | Blocked | N |
+4. Recent activity (last 5 actions from learning/feedback.jsonl)
+
+## Roadmap View
+1. Scan workspace/tasks/ for all tasks
+2. Categorize:
+   - **Now**: in_progress tasks
+   - **Next**: planning tasks (ranked by RICE)
+   - **Later**: ideas with no task yet
+3. Show milestones (if configured)
+4. Show dependency graph
+5. Highlight blocked items
+
+## Health Score
+Calculate from:
+| Dimension | Weight | Source |
+| EVA quality trend | 25% | tail .qoder/learning/eval-history.jsonl (avg score_pct, target >=80) |
+| Index freshness | 20% | data/index/.index-meta.json last_sync (<=7 days) + verify checks |
+| On-time delivery | 20% | task completion dates |
+| Sync health | 15% | python .qoder/scripts/team_sync.py status (ahead/behind/dirty) |
+| Pipeline flow | 10% | PRD->Spec->Code conversion |
+| Learning growth | 10% | learning/patterns/ entries |
+
+Score: >=4 healthy / 3-4 at risk / <3 needs attention
+Also surface: 索引新鲜度天数、最近 3 次 EVA 分数、未同步产出数。

package/templates/qoder/commands/wl-task.md

@@ -0,0 +1,71 @@
+---
+name: wl-task
+description: "Manage tasks with built-in RICE prioritization. Actions: create/list/start/finish/archive/rank"
+argument-hint: "[action] [task name or ID]"
+auto-approve: true
+allowed-tools: [Read, Glob, Grep, Bash, Write, Edit]
+---
+
+# /wl-task - Task Management with Prioritization
+
+User input: $ARGUMENTS
+
+## Actions
+- create [name]  - Create new task (`python .qoder/scripts/task.py create "<title>"`)
+- list           - List all tasks (`task.py list [--mine] [--status s]`)
+- show [ID]      - Show task details (`task.py show <name>`)
+- start [ID]     - Start task (`task.py start <name>`, planning -> in_progress)
+- finish         - Finish current task (`task.py finish`, -> completed)
+- archive [ID]   - Archive completed task (`task.py archive <name>`)
+- rank           - Re-prioritize all tasks using RICE (AI-driven, see below)
+
+## Built-in Prioritization (no /prioritize needed)
+When creating multiple tasks or running "rank":
+1. Run `task.py list` to get all pending tasks
+2. Score each task: R(Reach) x I(Impact) x C(Confidence) / E(Effort)
+3. Auto-sort by RICE score and show ranked list with scores
+4. Ask: adjust any scores?
+5. Write the confirmed score into each task.json as `"rice": <score>` (Edit tool)
+
+## Storage
+Tasks stored in: workspace/tasks/{MM-DD-slug}/
+Each task: task.json + prd.md + implement.jsonl + check.jsonl
+Archived tasks move to: .qoder/archive/{YYYY-MM}/{MM-DD-slug}/
+
+---
+
+## Flow
+
+### create
+1. Run `python .qoder/scripts/task.py create "<title>" [--assignee dev] [--priority P0-P3]`
+2. Task directory is created with task.json (status=planning) and prd.md template
+3. If user provides description -> fill it into prd.md
+4. If multiple tasks pending -> auto-suggest ranking
+5. Auto-sync: `python .qoder/scripts/team_sync.py push`
+
+### list
+Run `python .qoder/scripts/task.py list` and show as table:
+| ID | Name | Status | Priority | RICE | Assignee | Updated |
+
+### rank
+1. List all planning/pending tasks
+2. For each: estimate R, I, C, E
+3. Calculate RICE scores
+4. Show ranked result
+5. Ask: confirm order? Then persist scores to task.json
+
+### start
+1. Run `python .qoder/scripts/task.py start <name>`
+2. Status: planning -> in_progress, task becomes the active task
+3. Load PRD + context, ready for /wl-spec or /wl-code
+
+### finish
+1. Run `python .qoder/scripts/task.py finish`
+2. Status -> completed, active task pointer cleared
+3. Record to learning system
+4. Auto-sync: `python .qoder/scripts/team_sync.py push`
+
+### archive
+1. Run `python .qoder/scripts/task.py archive <name>`
+2. Moves task to .qoder/archive/{YYYY-MM}/
+3. Record completion metrics

package/templates/qoder/commands/wl-test.md

@@ -0,0 +1,42 @@
+---
+name: wl-test
+description: "Generate unit tests for implemented code. Needs confirmation."
+argument-hint: "[class-name or file-path]"
+auto-approve: false
+allowed-tools: [Read, Glob, Grep, Bash, Write, Edit]
+---
+
+# /wl-test - Generate Tests
+
+User input: $ARGUMENTS
+
+## Step 1: Find Target Code
+
+If class-name provided, search for it:
+  python .qoder/scripts/search_index.py <class-name>
+If no argument, use current task's implemented code.
+
+## Step 2: Read Code + Spec
+
+Read the implementation file and related spec.
+Understand what needs to be tested.
+
+## Step 3: Generate Tests
+
+Read test-generator skill for testing workflow.
+Generate tests covering:
+1. Happy path for each function
+2. Edge cases (null, empty, boundary values)
+3. Error handling
+4. Integration points
+
+## Step 4: Self-Check
+
+- [ ] Tests are runnable
+- [ ] Cover main scenarios
+- [ ] Use proper assertions
+- [ ] Follow team test conventions
+
+## Step 5: Report
+
+Tell user what tests were generated and suggest /wl-commit next.

package/templates/qoder/config.toml

@@ -0,0 +1,5 @@
+# Qoder CLI 项目级配置
+# 文档: https://docs.qoder.com/zh/cli/permissions
+
+# 权限模式: auto = 自动批准 (不弹确认)
+approval_mode = "auto"