speccrew 0.6.68 → 0.7.0

package/.speccrew/skills/speccrew-knowledge-techs-generate-conventions/SKILL.md

@@ -1,21 +1,13 @@
 ---
 name: speccrew-knowledge-techs-generate-conventions
-description: Generate technology convention documents (INDEX, tech-stack, architecture, conventions-*) for a specific platform. Extracts tech stack, architecture, and development conventions from configuration files and source code. Split from techs-generate for parallel execution with ui-style worker.
+description: Generate technology convention documents (INDEX, tech-stack, architecture, conventions-*) for a specific platform using XML Block workflow. Extracts tech stack, architecture, and development conventions from configuration files and source code. Split from techs-generate for parallel execution with ui-style worker.
 tools: Read, Write, Glob, Grep
 ---
 
-# Stage 2: Generate Platform Convention Documents
+# Stage 2: Generate Platform Convention Documents (XML Block Workflow)
 
 Generate comprehensive convention documentation for a specific platform by analyzing its configuration files and source code structure. This skill focuses on conventions documents only; UI style analysis is handled by the separate techs-generate-ui-style worker.
 
-## Language Adaptation
-
-**CRITICAL**: Generate all content in the language specified by the `language` parameter.
-
-- `language: "zh"` → Generate all content in 中文
-- `language: "en"` → Generate all content in English
-- Other languages → Use the specified language
-
 ## Trigger Scenarios
 
 - "Generate convention documents for {platform}"
@@ -28,31 +20,42 @@ Generate comprehensive convention documentation for a specific platform by analy
 
 Worker Agent (speccrew-task-worker)
 
-## Input
-
-- `platform_id`: Platform identifier (e.g., "web-react", "backend-nestjs")
-- `platform_type`: Platform type (web, mobile, backend, desktop)
-- `framework`: Primary framework (react, nestjs, flutter, etc.)
-- `source_path`: Platform source directory
-- `config_files`: List of configuration file paths
-- `convention_files`: List of convention file paths (eslint, prettier, etc.)
-- `output_path`: Output directory for generated documents (e.g., `speccrew-workspace/knowledges/techs/{platform_id}/`)
-- `language`: Target language (e.g., "zh", "en") - **REQUIRED**
-- `completed_dir`: (Optional) Directory for analysis coverage report output. If provided, the analysis JSON will be written here instead of the knowledges directory
+## Input Variables
+
+| Variable | Type | Description | Example |
+|----------|------|-------------|---------|
+| `${platform_id}` | string | Platform identifier | `"web-react"`, `"backend-nestjs"` |
+| `${platform_type}` | string | Platform type | `web`, `mobile`, `backend`, `desktop`, `api` |
+| `${framework}` | string | Primary framework | `react`, `nestjs`, `flutter`, etc. |
+| `${source_path}` | string | Platform source directory | `"frontend-web"` |
+| `${config_files}` | array | List of configuration file paths | `["package.json", "tsconfig.json"]` |
+| `${convention_files}` | array | List of convention file paths | `[".eslintrc.js", ".prettierrc"]` |
+| `${output_path}` | string | Output directory for generated documents | `speccrew-workspace/knowledges/techs/{platform_id}/` |
+| `${language}` | string | Target language for generated content | `"zh"`, `"en"` |
+| `${completed_dir}` | string | (Optional) Directory for analysis coverage report output | `speccrew-workspace/iterations/...` |
+
+## Output Variables
+
+| Variable | Type | Description |
+|----------|------|-------------|
+| `${status}` | string | Generation status: `"success"`, `"partial"`, or `"failed"` |
+| `${documents_generated}` | array | List of generated document filenames |
+| `${analysis_file}` | string | Path to the analysis coverage report |
+| `${message}` | string | Summary message for status update |
 
 ## Output
 
-Generate the following documents in `{output_path}/`:
+Generate the following documents in `${output_path}/`:
 
 ```
-{output_path}/
+${output_path}/
 ├── INDEX.md                    # Platform technology index (Required)
 ├── tech-stack.md              # Technology stack details (Required)
 ├── architecture.md            # Architecture conventions (Required)
 ├── conventions-design.md      # Design conventions (Required)
 ├── conventions-dev.md         # Development conventions (Required)
-├── conventions-unit-test.md        # Unit testing conventions (Required)
-├── conventions-system-test.md      # System testing conventions (Required)
+├── conventions-unit-test.md   # Unit testing conventions (Required)
+├── conventions-system-test.md # System testing conventions (Required)
 ├── conventions-build.md       # Build & Deployment conventions (Required)
 └── conventions-data.md        # Data conventions (Optional)
 ```
@@ -67,582 +70,58 @@ Generate the following documents in `{output_path}/`:
 | `desktop` | All 8 docs | conventions-data.md | **Default No** - Based on actual tech stack |
 | `api` | All 8 docs | conventions-data.md | **Conditional** - Based on whether data layer exists |
 
-### Decision Logic for conventions-data.md
-
-**Step 1: Check Platform Type**
-- If `backend` → **Generate** (always)
-- If `web`/`mobile`/`desktop`/`api` → Proceed to Step 2
-
-**Step 2: Detect Data Layer (for non-backend platforms)**
-
-Check configuration files for data layer indicators:
-
-| Indicator | Technology | Action |
-|-----------|------------|--------|
-| `prisma` in package.json dependencies | Prisma ORM | Generate conventions-data.md |
-| `typeorm` in package.json dependencies | TypeORM | Generate conventions-data.md |
-| `sequelize` in package.json dependencies | Sequelize | Generate conventions-data.md |
-| `mongoose` in package.json dependencies | Mongoose | Generate conventions-data.md |
-| `drizzle-orm` in package.json dependencies | Drizzle ORM | Generate conventions-data.md |
-| `firebase` / `@react-native-firebase` | Firebase | Generate conventions-data.md (lightweight) |
-| `sqlite` / `realm` / `@realm/react` | SQLite/Realm | Generate conventions-data.md (lightweight) |
-| `core-data` in iOS project | Core Data | Generate conventions-data.md |
-| `room` in Android project | Room Persistence | Generate conventions-data.md |
-| None detected | - | **Skip** conventions-data.md |
-
-**Step 3: Report Decision**
-```
-Platform: {platform_id}
-Type: {platform_type}
-Framework: {framework}
-Data Layer Detected: {yes/no/technology}
-Generate conventions-data.md: {yes/no}
-```
-
-## Workflow
-
-```mermaid
-flowchart TD
-    Start --> Step0[Step 0: Read Document Templates]
-    Step0 --> Step1[Step 1: Read Configuration Files]
-    Step1 --> Step2[Step 2: Extract Technology Stack]
-    Step2 --> Step3[Step 3: Analyze Conventions]
-    Step3 --> Step4[Step 4: Generate Documents]
-    Step4 --> Step5[Step 5: Write Output Files]
-    Step5 --> Step6[Step 6: Generate Analysis Coverage Report]
-    Step6 --> Step7[Step 7: Report Results]
-    Step7 --> End
-```
-
-### Step 0: Read Document Templates
-
-Before processing, read all template files to understand the required content structure for each document type:
-- **Read**: `../speccrew-knowledge-techs-generate/templates/INDEX-TEMPLATE.md` - Platform overview and navigation hub structure
-- **Read**: `../speccrew-knowledge-techs-generate/templates/TECH-STACK-TEMPLATE.md` - Technology stack details structure
-- **Read**: `../speccrew-knowledge-techs-generate/templates/ARCHITECTURE-TEMPLATE.md` - Architecture patterns and conventions structure
-- **Read**: `../speccrew-knowledge-techs-generate/templates/CONVENTIONS-DESIGN-TEMPLATE.md` - Design principles and patterns structure
-- **Read**: `../speccrew-knowledge-techs-generate/templates/CONVENTIONS-DEV-TEMPLATE.md` - Development conventions structure
-- **Read**: `../speccrew-knowledge-techs-generate/templates/CONVENTIONS-UNIT-TEST-TEMPLATE.md` - Unit testing conventions structure
-- **Read**: `../speccrew-knowledge-techs-generate/templates/CONVENTIONS-SYSTEM-TEST-TEMPLATE.md` - System testing conventions structure
-- **Read**: `../speccrew-knowledge-techs-generate/templates/CONVENTIONS-BUILD-TEMPLATE.md` - Build and deployment conventions structure
-- **Read**: `../speccrew-knowledge-techs-generate/templates/CONVENTIONS-DATA-TEMPLATE.md` - Data layer conventions structure (if applicable)
-- **Purpose**: Understand each template's chapters and example content requirements
-- **Key principle**: Extract information from source code according to template section requirements
-
-### Step 1: Read Configuration Files
-
-Read and parse all configuration files for the platform:
-
-**Primary Config Files:**
-- package.json / pom.xml / requirements.txt / pubspec.yaml / go.mod
-- tsconfig.json / jsconfig.json
-- Build config: vite.config.* / webpack.config.* / next.config.* / nest-cli.json
-
-**Convention Files:**
-- ESLint: .eslintrc.* / eslint.config.*
-- Prettier: .prettierrc.* / prettier.config.*
-- Testing: jest.config.* / vitest.config.* / pytest.ini
-- Git: .gitignore, .gitattributes
-
-### Step 2: Extract Technology Stack
-
-Parse configuration files to extract:
-
-**Core Framework:**
-- Name and version from dependencies
-- Primary language (TypeScript, JavaScript, Dart, etc.)
-
-**Dependencies:**
-- Production dependencies (grouped by purpose)
-- Development dependencies
-- Key library versions
-
-**Build Tools:**
-- Bundler (Vite, Webpack, Rollup)
-- Transpiler (TypeScript, Babel)
-- Task runner (npm scripts, Gradle, Maven)
-
-**Example Extraction:**
-```json
-{
-  "framework": "React",
-  "framework_version": "18.2.0",
-  "language": "TypeScript",
-  "language_version": "5.3.0",
-  "build_tool": "Vite 5.0.0",
-  "key_dependencies": [
-    { "name": "react-router-dom", "version": "6.20.0", "purpose": "Routing" },
-    { "name": "zustand", "version": "4.4.0", "purpose": "State Management" }
-  ]
-}
-```
-
-### Step 3: Analyze Conventions
-
-1. **Read Configuration**:
-   - Read `speccrew-workspace/docs/rules/mermaid-rule.md` - Get Mermaid diagram compatibility guidelines
-
-2. **Extract conventions from configuration files:**
-
-**From ESLint Config:**
-- Enabled rules
-- Code style preferences
-- Import/export patterns
-
-**From Prettier Config:**
-- Formatting rules (semi, quotes, tabWidth)
-- Print width
-
-**From Project Structure:**
-- Directory conventions (src/, components/, utils/)
-- File naming patterns
-- Module organization
-
-3. **Apply Mermaid Rules**:
-   - Follow compatibility guidelines from `mermaid-rule.md`
-   - See: [Mermaid Diagram Guide](#mermaid-diagram-guide)
-
-### Domain-Specific Convention Extraction (MANDATORY)
-
-In addition to the general conventions above, you MUST actively search for and extract the following domain-specific topics. These are critical for downstream Agents (solution, design, development, testing).
-
-**For Frontend Platforms (web-vue, mobile-uniapp, etc.):**
-
-| Topic | What to Search For | Where to Look |
-|-------|-------------------|---------------|
-| **i18n/Internationalization** | i18n framework config, locale files, translation key patterns | `locales/`, `i18n/`, `lang/`, package.json deps |
-| **Authorization & Permissions** | Permission directives (`v-hasPermi`), route guards, permission stores | `permission/`, `router/`, `store/`, `utils/auth` |
-| **Menu Registration** | Menu config, dynamic menu loading, menu-to-route mapping | `router/`, `store/`, `layout/`, API calls for menus |
-| **Data Dictionary** | Dict components (`DictTag`), dict stores, dict API calls | `components/Dict`, `utils/dict`, `store/` |
-| **Logging** | Error reporting service, console policy, error boundaries | `utils/log`, `plugins/sentry`, error handling config |
-| **API Request Layer** | Axios/fetch instance, interceptors, token refresh, base URL config | `utils/request`, `api/`, `config/`, `interceptors/` |
-| **Data Validation** | Form validation rules, custom validators, validation timing | `utils/validate`, form schemas, component props validation |
-| **File Upload** | Upload components, upload API calls, file size limits | `components/Upload`, `api/file`, `utils/upload` |
-
-**For Backend Platforms (backend-spring, etc.):**
-
-| Topic | What to Search For | Where to Look |
-|-------|-------------------|---------------|
-| **Authorization & Permissions** | `@PreAuthorize`, `@DataPermission`, security config, permission enums | Security config, controller annotations, framework modules |
-| **Data Dictionary** | Dict entity, dict service, dict cache, dict enum patterns | `dict/`, `system/` module, enum classes |
-| **Multi-tenancy** | Tenant interceptor/plugin, tenant column, tenant context, `@TenantIgnore` | MyBatis plugins, framework config, base entity |
-| **Backend i18n** | messages.properties, MessageSource, i18n config, ValidationMessages | `resources/i18n/`, `messages*.properties`, i18n config beans |
-| **Logging** | Logger usage, log config (logback.xml/log4j2), operation log annotation, audit trail | `logback*.xml`, `log4j2*.xml`, `@OperateLog`, `operatelog/` module |
-| **Exception Handling** | GlobalExceptionHandler, business exceptions, error codes, error response format | `handler/`, `exception/`, `enums/ErrorCode`, `GlobalExceptionHandler` |
-| **Caching** | @Cacheable, RedisTemplate, cache key patterns, cache config | `cache/`, `redis/`, `CacheConfig`, `@Cacheable` annotations |
-| **Data Validation** | @Valid, @Validated, custom validators, validation groups | DTO classes, `validator/`, `@NotNull/@Size` patterns |
-| **Scheduled Jobs** | @Scheduled, Quartz config, XXL-Job handler, cron expressions | `job/`, `task/`, `schedule/`, `@Scheduled` methods |
-| **File Storage** | FileService, upload API, OSS/S3 config, file path patterns | `file/`, `infra/file/`, `FileClient`, storage config |
-
-**If a topic is not found in the source code**, explicitly state "Not applicable" in the corresponding template section. Do NOT leave the section empty or skip it silently.
-
-### Analysis Tracking (MANDATORY)
-
-During Step 3, you MUST maintain an internal tracking record for each topic you search. For every topic in the tables above:
-
-1. **Search** the source code using the "Where to Look" paths
-2. **Record** the result:
-   - `found` — Topic implementation found, relevant files identified
-   - `not_found` — Searched all suggested paths, no implementation exists
-   - `partial` — Some aspects found but incomplete
-3. **List** all files you actually read/analyzed for that topic
-
-This tracking data will be used in Step 6 to generate the analysis coverage report. Do NOT skip any topic — if a topic is not applicable to this platform type, record it as `not_found` with a note.
-
-**Topic Checklist by Platform Type:**
-
-**Frontend Topics (web, mobile, desktop):**
-1. i18n / Internationalization
-2. Authorization & Permissions
-3. Menu Registration & Routing
-4. Data Dictionary Usage
-5. Logging & Error Reporting
-6. API Request Layer (Axios/fetch)
-7. Data Validation
-8. File Upload & Storage
-9. UI Style System is handled by the separate techs-generate-ui-style worker
-
-**Backend Topics (backend):**
-1. Backend Internationalization
-2. Authorization & Permissions (annotations, data permission)
-3. Data Dictionary Management
-4. Logging & Audit Trail
-5. Exception Handling & Error Codes
-6. Caching Strategy
-7. Data Validation (JSR 380, custom validators)
-8. Scheduled Jobs & Task Scheduling
-9. File Storage
-10. Multi-tenancy
-
-### Step 4: Generate Documents (MANDATORY: Copy Template + Fill)
-
-**CRITICAL**: This step MUST follow the template fill workflow - copy template first, then fill sections.
-
-1. **For Each Document, Follow This Workflow**:
-
-   **Step 4.1: Copy Template File**
-   - Copy the corresponding template file to the output path:
-     - `../speccrew-knowledge-techs-generate/templates/INDEX-TEMPLATE.md` → `{output_path}/INDEX.md`
-     - `../speccrew-knowledge-techs-generate/templates/TECH-STACK-TEMPLATE.md` → `{output_path}/tech-stack.md`
-     - `../speccrew-knowledge-techs-generate/templates/ARCHITECTURE-TEMPLATE.md` → `{output_path}/architecture.md`
-     - `../speccrew-knowledge-techs-generate/templates/CONVENTIONS-DESIGN-TEMPLATE.md` → `{output_path}/conventions-design.md`
-     - `../speccrew-knowledge-techs-generate/templates/CONVENTIONS-DEV-TEMPLATE.md` → `{output_path}/conventions-dev.md`
-     - `../speccrew-knowledge-techs-generate/templates/CONVENTIONS-UNIT-TEST-TEMPLATE.md` → `{output_path}/conventions-unit-test.md`
-     - `../speccrew-knowledge-techs-generate/templates/CONVENTIONS-SYSTEM-TEST-TEMPLATE.md` → `{output_path}/conventions-system-test.md`
-     - `../speccrew-knowledge-techs-generate/templates/CONVENTIONS-BUILD-TEMPLATE.md` → `{output_path}/conventions-build.md`
-     - `../speccrew-knowledge-techs-generate/templates/CONVENTIONS-DATA-TEMPLATE.md` → `{output_path}/conventions-data.md` (if applicable)
-
-   **Step 4.2: Fill Template Sections with search_replace**
-   - Use `search_replace` tool to fill each section of the template
-   - Replace placeholder content with actual analyzed data
-     - Follow [Document Structure Standard](#document-structure-standard)
-   - Apply [Source Traceability Requirements](#source-traceability-requirements)
-
-   **MANDATORY RULES**:
-   - **Do NOT use create_file to rewrite the entire document**
-   - **Do NOT delete or skip any template section**
-   - Only replace the placeholder content within each section
-   - Preserve all template section headers and structure
-
-2. **Document Generation Order**:
-   - Generate: INDEX.md, tech-stack.md, architecture.md, conventions-design.md, conventions-dev.md, conventions-unit-test.md, conventions-system-test.md, conventions-build.md, conventions-data.md (if applicable)
-
-3. **UI Design Conventions Reference in conventions-design.md**:
-   In conventions-design.md, ALWAYS include UI reference section:
-   ```markdown
-   ## UI Design Conventions
-
-   Refer to [UI Style Guide](ui-style/ui-style-guide.md) for design system details.
-   Note: UI style documents are generated by the separate ui-style worker.
-   ```
-
-### Step 5: Write Output Files
-
-Create output directory if not exists, then write all generated documents.
-
-### Step 6: Generate Analysis Coverage Report (MANDATORY)
-
-After completing all document generation, you MUST create an analysis coverage report as a JSON file.
-
-**Output file**: `{completed_dir}/{platform_id}.analysis-conventions.json`
-
-Where `{completed_dir}` is the directory passed via the `completed_dir` parameter (if provided). If `completed_dir` is not provided, output to the platform's knowledges directory.
-
-**Report Format**:
-
-```json
-{
-  "platform_id": "{platform_id}",
-  "platform_type": "{platform_type}",
-  "worker_type": "conventions",
-  "analyzed_at": "{ISO 8601 timestamp}",
-  "topics": {
-    "i18n": {
-      "status": "found",
-      "files_analyzed": ["src/i18n/index.ts", "locales/zh-CN.ts"],
-      "notes": "Vue I18n with 2 locale files"
-    },
-    "authorization": {
-      "status": "found",
-      "files_analyzed": ["src/permission/index.ts", "src/router/guard.ts"],
-      "notes": "RBAC with route guards and v-hasPermi directive"
-    },
-    "data_dictionary": {
-      "status": "not_found",
-      "files_analyzed": [],
-      "notes": "No dictionary implementation found in suggested paths"
-    }
-  },
-  "config_files_analyzed": [
-    "package.json",
-    "vite.config.ts",
-    "tsconfig.json"
-  ],
-  "source_dirs_scanned": [
-    "src/components/",
-    "src/views/",
-    "src/utils/",
-    "src/store/"
-  ],
-  "documents_generated": [
-    "INDEX.md",
-    "tech-stack.md",
-    "architecture.md",
-    "conventions-dev.md",
-    "conventions-design.md",
-    "conventions-unit-test.md",
-    "conventions-build.md"
-  ],
-  "coverage_summary": {
-    "topics_found": 7,
-    "topics_partial": 1,
-    "topics_not_found": 1,
-    "topics_total": 9,
-    "coverage_percent": 78
-  }
-}
-```
-
-**Rules**:
-- Every topic from the Topic Checklist in Step 3 MUST appear in the `topics` object
-- `files_analyzed` MUST list the actual file paths you read (relative to source_path)
-- `status` MUST be one of: `found`, `not_found`, `partial`
-- `coverage_percent` = (topics_found + topics_partial) / topics_total * 100, rounded to integer
-- `documents_generated` MUST list all .md files actually created
-- Use `create_file` to write this JSON file (this is the ONE exception where create_file is allowed — for JSON output files)
-
-### Step 7: Report Results
-
-```
-Platform Convention Documents Generated: {{platform_id}}
-- INDEX.md: ✓
-- tech-stack.md: ✓
-- architecture.md: ✓
-- conventions-design.md: ✓
-- conventions-dev.md: ✓
-- conventions-unit-test.md: ✓
-- conventions-system-test.md: ✓
-- conventions-build.md: ✓
-- conventions-data.md: ✓ (or skipped if not applicable)
-- {{platform_id}}.analysis-conventions.json: ✓ (analysis coverage report)
-- Output Directory: {{output_path}}
-- Analysis Report: {{completed_dir}}/{{platform_id}}.analysis-conventions.json (or knowledges dir if completed_dir not provided)
-```
-
-**Completion Marker File**:
-
-Create a done file to signal completion:
-```json
-{
-  "platform_id": "{platform_id}",
-  "worker_type": "conventions",
-  "status": "completed",
-  "documents_generated": ["INDEX.md", "tech-stack.md", "architecture.md", "conventions-design.md", "conventions-dev.md", "conventions-unit-test.md", "conventions-system-test.md", "conventions-build.md"],
-  "analysis_file": "{platform_id}.analysis-conventions.json",
-  "completed_at": "{ISO timestamp}"
-}
-```
-
----
-
-## Reference Guides
-
-### Document Structure Standard
-
-All generated documents must follow this structure:
-
-```markdown
-# {{platform_name}} {{document_type}}
-
-**Files Referenced in This Document**
-
-{{source_files}}
-
-> **Target Audience**: devcrew-designer-{{platform_id}}, devcrew-dev-{{platform_id}}, devcrew-test-{{platform_id}}
-
-## Table of Contents
-1. [Introduction](#introduction)
-2. [Project Structure](#project-structure)
-3. [Core Components](#core-components)
-4. [Architecture Overview](#architecture-overview)
-5. [Detailed Component Analysis](#detailed-component-analysis)
-6. [Dependency Analysis](#dependency-analysis)
-7. [Performance Considerations](#performance-considerations)
-8. [Troubleshooting Guide](#troubleshooting-guide)
-9. [Conclusion](#conclusion)
-10. [Appendix](#appendix)
-
-... content sections ...
-
-**Section Source**
-- [file.ext](../../../../path/to/file#Lstart-Lend)
-```
-
-### Source Traceability Requirements
-
-**CRITICAL: All source file links MUST use RELATIVE PATHS. Absolute paths and `file://` protocol are STRICTLY FORBIDDEN.**
-
-**FORBIDDEN:**
-- Do NOT use absolute paths (e.g., `d:/dev/ruoyi-vue-pro/...`)
-- Do NOT use `file://` protocol (e.g., `file://d:/dev/...`)
-- Do NOT hardcode machine-specific paths
-- ALWAYS use relative paths calculated from the document's location
-
-**Dynamic Relative Path Calculation:**
+## AgentFlow Definition
 
-Documents are located at: `speccrew-workspace/knowledges/techs/{platform_id}/{document}.md`
-This is 4 levels deep from the project root:
-  - speccrew-workspace (1)
-  - knowledges (2)
-  - techs (3)
-  - {platform_id} (4)
+<!-- @agentflow: workflow.agentflow.xml -->
 
-Therefore, to reference a source file from the project root, use `../../../../` as the prefix.
+> **REQUIRED**: Before executing this workflow, read the XML workflow specification: `speccrew-workspace/docs/rules/agentflow-spec.md`
 
-Example calculation:
-  - Document: `speccrew-workspace/knowledges/techs/backend-spring/architecture.md`
-  - Source file: `yudao-server/src/main/java/.../YudaoServerApplication.java`
-  - Relative path: `../../../../yudao-server/src/main/java/.../YudaoServerApplication.java`
 
-For root INDEX.md (one level less deep):
-  - Document: `speccrew-workspace/knowledges/techs/INDEX.md`
-  - Prefix: `../../../` (3 levels)
 
-**1. File Reference Block**
+## Constraints
 
-Place at the beginning of each document (using table format for VS Code preview compatibility):
+1. **DO NOT create temporary scripts, batch files, or workaround code files (`.py`, `.bat`, `.sh`, `.ps1`, etc.)** under any circumstances
+2. **DO NOT analyze files outside the specified `${source_path}`**
+3. **All content MUST be in the language specified by `${language}`**
+4. **Use `search_replace` for section filling, NEVER rewrite entire document**
+5. **Mermaid diagrams MUST follow the rules in `mermaid-rule.md`**
+6. **All links MUST use relative paths, NEVER `file://` protocol**
+7. **Write each document to file immediately after generation - DO NOT accumulate all documents in memory**
+8. **DO NOT create done marker file until ALL required documents have been verified to exist on disk**
 
-```markdown
-**Files Referenced in This Document**
-
-| # | File | Source |
-|---|------|--------|
-| 1 | package.json | [View](../../../../yudao-ui/yudao-ui-admin-vue3/package.json) |
-| 2 | tsconfig.json | [View](../../../../yudao-ui/yudao-ui-admin-vue3/tsconfig.json) |
-```
-
-> **Note**: Table format is used instead of list format because LLM-generated content may merge list items into a single line, breaking link rendering in VS Code preview.
-
-**2. Diagram Source Annotation**
-
-After each Mermaid diagram:
-
-```markdown
-**Diagram Source**
-- [file-name.ext](../../../../yudao-server/src/main/java/...#L10-L50)
-```
-
-**3. Section Source Annotation**
-
-At the end of each major section:
-
-```markdown
-**Section Source**
-- [file-name.ext](../../../../yudao-server/src/main/java/...#L10-L50)
-```
-
-For generic guidance sections without specific file references:
-
-```markdown
-[This section provides general guidance, no specific file reference required]
-```
-
-### Mermaid Diagram Guide
-
-When generating Mermaid diagrams, follow these compatibility guidelines:
-
-**Key Requirements:**
-- Use only basic node definitions: `A[text content]`
-- No HTML tags (e.g., `<br/>`)
-- No nested subgraphs
-- No `direction` keyword
-- No `style` definitions
-- Use standard `graph TB/LR` syntax only
-
-**Diagram Types:**
-
-| Diagram Type | Use Case | Example Scenario |
-|--------------|----------|------------------|
-| `graph TB/LR` | Structure & Dependency | Module relationships, component hierarchy |
-| `flowchart TD` | Business Logic Flow | Request processing, decision trees |
-| `sequenceDiagram` | Interaction Flow | API calls, service communication |
-| `classDiagram` | Class Structure | Entity relationships, inheritance |
-| `erDiagram` | Database Schema | Table relationships, data model |
-| `stateDiagram-v2` | State Machine | Order status, workflow states |
-
-### Quality Requirements
-
-#### Be Specific
-
-Extract actual values from config files:
-- ✓ "React 18.2.0" (from package.json)
-- ✗ "React (version varies)"
-
-#### Be Concise
-
-Focus on actionable conventions:
-- ✓ "Use PascalCase for component files: UserProfile.tsx"
-- ✗ "There are many naming conventions to consider..."
-
-#### Include Examples
-
-Wherever possible, include concrete examples:
-```markdown
-### Component Naming
-- ✓ UserProfile.tsx
-- ✓ OrderList.tsx
-- ✗ userProfile.tsx
-- ✗ order-list.tsx
-```
-
-### Error Handling
-
-**Template Not Found:**
-- If a template file is missing, report error and skip that document
-- Continue with other documents
-
-**Configuration File Missing:**
-- If expected config file not found, note in analysis report
-- Use defaults or skip related sections
-
-**Source Code Not Found:**
-- If source directory structure doesn't match expectations
-- Document actual structure found
-- Mark topics as "not_found" in analysis report
+## Task Completion Report
 
----
+When the task is complete, report the following:
 
-## Task Completion Report
+**Status:** `success` | `partial` | `failed`
 
-Upon completion, output the following structured report:
-
-```json
-{
-  "status": "success | partial | failed",
-  "skill": "speccrew-knowledge-techs-generate-conventions",
-  "output_files": [
-    "{output_path}/INDEX.md",
-    "{output_path}/tech-stack.md",
-    "{output_path}/architecture.md",
-    "{output_path}/conventions-design.md",
-    "{output_path}/conventions-dev.md",
-    "{output_path}/conventions-unit-test.md",
-    "{output_path}/conventions-system-test.md",
-    "{output_path}/conventions-build.md",
-    "{output_path}/conventions-data.md",
-    "{completed_dir}/{platform_id}.analysis-conventions.json"
-  ],
-  "summary": "Convention documents generated for {platform_id} with {coverage_percent}% topic coverage",
-  "metrics": {
-    "convention_categories": 0,
-    "rules_documented": 0,
-    "code_examples_included": 0
-  },
-  "errors": [],
-  "next_steps": [
-    "Run speccrew-knowledge-techs-generate-ui-style for frontend platforms",
-    "Review analysis-conventions.json for topic coverage gaps"
-  ]
-}
-```
+**Summary:**
+- Platform: `${platform_id}`
+- Type: `${platform_type}`
+- Framework: `${framework}`
+- Documents generated: 8 required + (1 optional if data layer detected)
 
----
+**Files Generated:**
+- `${output_path}/INDEX.md` - Platform technology index
+- `${output_path}/tech-stack.md` - Technology stack details
+- `${output_path}/architecture.md` - Architecture conventions
+- `${output_path}/conventions-design.md` - Design conventions
+- `${output_path}/conventions-dev.md` - Development conventions
+- `${output_path}/conventions-unit-test.md` - Unit testing conventions
+- `${output_path}/conventions-system-test.md` - System testing conventions
+- `${output_path}/conventions-build.md` - Build and deployment conventions
+- `${output_path}/conventions-data.md` - Data conventions (optional)
+- `${completed_dir}/${platform_id}.analysis-conventions.json` - Analysis coverage report
 
 ## Checklist
 
 ### Pre-Generation
+- [ ] All template files read successfully
 - [ ] All configuration files read and parsed
 - [ ] Technology stack extracted accurately
 - [ ] Conventions analyzed from config files
 
-### Document Generation Decision
-- [ ] Platform type identified (web/mobile/backend/desktop/api)
-- [ ] Data layer detection completed for non-backend platforms
-- [ ] Decision made on whether to generate conventions-data.md
-  - [ ] Backend platform → Always generate
-  - [ ] Other platforms → Generate only if data layer detected
-
-### Required Documents (All Platforms)
+### Document Generation
 - [ ] INDEX.md generated with navigation
 - [ ] tech-stack.md generated with dependency tables
 - [ ] architecture.md generated with platform-specific patterns
@@ -651,8 +130,6 @@ Upon completion, output the following structured report:
 - [ ] conventions-unit-test.md generated with unit testing requirements
 - [ ] conventions-system-test.md generated with system testing requirements
 - [ ] conventions-build.md generated with build and deployment conventions
-
-### Optional Document
 - [ ] conventions-data.md generated (only if applicable per platform type mapping)
 
 ### Quality Checks
@@ -661,6 +138,17 @@ Upon completion, output the following structured report:
 - [ ] **Source traceability**: Diagram Source annotations added after each Mermaid diagram
 - [ ] **Source traceability**: Section Source annotations added at end of major sections
 - [ ] **Mermaid compatibility**: No `style`, `direction`, `<br/>`, or nested subgraphs
-- [ ] **Document completeness**: Verify all 8 required documents exist (INDEX.md, tech-stack.md, architecture.md, conventions-design.md, conventions-dev.md, conventions-unit-test.md, conventions-system-test.md, conventions-build.md)
-- [ ] **Analysis Coverage Report**: `{platform_id}.analysis-conventions.json` generated with all topics tracked
-- [ ] Results reported with conventions-data.md generation status
+- [ ] **Document completeness**: All 8 required documents exist (verified in Step 3)
+- [ ] **Done marker integrity**: Done marker only created after verification
+- [ ] **Analysis Coverage Report**: `${platform_id}.analysis-conventions.json` generated
+
+## CONTINUOUS EXECUTION RULES
+
+This skill MUST execute continuously without user interruption:
+
+1. **All steps must complete in a single session** - from template reading to done marker creation
+2. **If context window is running low**: Save checkpoint and inform user - DO NOT create false done marker
+3. **No intermediate user confirmation required** between steps
+4. **Error handling**: On error, log details and stop - do not proceed with incomplete data
+5. **Verification gate**: Step 3 verification MUST pass before Step 4 done marker creation
+6. **Memory management**: Write each document immediately after generation, do not accumulate in memory