speccrew 0.7.74 → 0.7.76

package/.speccrew/skills/speccrew-feature-designer-orchestration/SKILL.md

@@ -1,117 +1,18 @@
 ---
 name: speccrew-feature-designer-orchestration
-version: 1.0.0
 description: Feature Designer 的核心编排技能，负责加载功能清单、协调功能分析、功能规格设计与 API 契约生成。处理单个功能（直接调用）和多个功能（Worker 分发）两种工作流路径。
 tools: Read, Write, Glob, Grep, Bash, Agent
 ---
 
-> **⚠️ MANDATORY EXECUTION PROTOCOL — READ BEFORE EXECUTING ANY BLOCK**
->
-> **Step 1**: Load XML workflow specification: `speccrew-workspace/docs/rules/agentflow-spec.md` — this defines all block types and action-to-tool mappings
->
-> **Step 2**: Execute this SKILL.md's XML workflow **block by block in document order**. For EVERY block, you MUST follow this 3-step cycle:
->
-> ```
-> 🏷️ Block [ID] (action=[action]) — [desc]
-> 🔧 Tool: [which IDE tool to call]
-> ✅ Result: [output or status]
-> ```
->
-> Action-to-tool mapping:
-> - `action="run-skill"` → Invoke via **Skill tool** (pass the `<field name="skill">` value EXACTLY)
-> - `action="run-script"` → Execute via **Terminal tool** (pass the `<field name="command">` value EXACTLY)
-> - `action="dispatch-to-worker"` → Create **Task** via **Task tool** for `speccrew-task-worker`
-> - `action="read-file"` → Read via **Read tool**
-> - `action="log"` → Output message directly
-> - `action="confirm"` → Present to user and wait for response
->
-> **Step 3**: Execute ALL blocks sequentially without pausing (only stop at explicit `<event action="confirm">` blocks)
+# Trigger Scenarios
 
-# Feature Designer Orchestration
-
-Feature Designer 的核心编排技能，负责：
-
-1. **Stage Gate Verification** - 验证 PRD 阶段已完成确认
-2. **Feature List Loading** - 从 .prd-feature-list.json 加载功能清单
-3. **Feature Analysis Orchestration** - 协调功能分析（单个直接调用，多个分发 Worker）
-4. **Feature Design Orchestration** - 协调功能规格设计
-5. **API Contract Generation** - 生成 API 契约文档
-6. **Joint Confirmation** - 汇总确认所有设计产出
-
-## Invocation Method
-
-**CRITICAL**: This skill is loaded directly by Feature Designer Agent — do NOT invoke via Worker Agent.
-
-```xml
-<block type="task" action="run-skill" desc="Feature Designer orchestration workflow">
-  <field name="skill">speccrew-feature-designer-orchestration</field>
-</block>
-```
-
-## Input Parameters
-
-| Parameter | Type | Required | Description |
-|-----------|------|----------|-------------|
-| `prd_path` | string | Yes | PRD 文档目录路径 |
-| `iteration_path` | string | Yes | 当前迭代目录路径 |
-| `workspace_root` | string | Yes | speccrew-workspace 根目录路径 |
-| `frontend_platforms` | array | No | 前端平台列表（从 techs-manifest 读取） |
-
-## Output
-
-- `status` - 执行状态 (success / partial / failed)
-- `feature_specs` - 生成的功能规格文件列表
-- `api_contracts` - 生成的 API 契约文件列表
-- `workflow_stage` - 当前工作流阶段状态
-- `next_agent` - 下一步建议的 Agent
-
----
+- Feature Designer Agent starts feature design workflow after PRD confirmation
+- PRD stage confirmed and feature list needs to be loaded
+- User requests feature specification design or API contract generation
 
 ## AgentFlow Definition
 
 <!-- @agentflow: SKILL.xml -->
 
----
-
-## CONTINUOUS EXECUTION RULES
-
-This skill MUST execute tasks continuously without unnecessary interruptions.
-
-### FORBIDDEN Interruptions
-
-1. DO NOT ask user "Should I continue?" after completing a subtask
-2. DO NOT suggest "Let me split this into batches" or "Let's do this in parts"
-3. DO NOT pause to list what you plan to do next — just do it
-4. DO NOT ask for confirmation before generating output files
-5. DO NOT warn about "large number of files" — proceed with generation
-6. DO NOT offer "Should I proceed with the remaining items?"
-
-### When to Pause (ONLY these cases)
-
-1. CHECKPOINT gates defined in workflow (user confirmation required by design)
-2. Ambiguous requirements that genuinely need clarification
-3. Unrecoverable errors that prevent further progress
-4. Security-sensitive operations (e.g., deleting existing files)
-
-### Worker Enforcement Rules
-
-This agent is an **orchestrator/dispatcher**. When multiple Features exist, it MUST delegate all skill execution to `speccrew-task-worker` agents.
-
-| Condition | Action | Tool |
-|-----------|--------|------|
-| 1 Feature | Direct skill invocation allowed | Skill tool |
-| 2+ Features | **MUST** dispatch Workers | speccrew-task-worker via Agent tool |
-
-### FORBIDDEN Actions (When Features ≥ 2)
-
-1. ❌ DO NOT invoke `speccrew-fd-feature-analyze` skill directly
-2. ❌ DO NOT invoke `speccrew-fd-feature-design` skill directly
-3. ❌ DO NOT invoke `speccrew-fd-api-contract` skill directly
-4. ❌ DO NOT generate `.feature-analysis.md` files yourself
-5. ❌ DO NOT generate `.feature-spec.md` files yourself
-6. ❌ DO NOT generate `.api-contract.md` files yourself
-7. ❌ DO NOT create any document content as fallback if worker fails
-
-### Name Lock Rule
-
-After Phase 2 Feature List is confirmed, feature_name is immutable. All Skills MUST use the exact parameter value for output filenames. Name translation or substitution is FORBIDDEN.
+> **REQUIRED**: Before executing this workflow, read the XML workflow specification: speccrew-workspace/docs/rules/agentflow-spec.md
+> Then read and execute the XML workflow in SKILL.xml block-by-block as the authoritative execution plan.

package/.speccrew/skills/speccrew-get-timestamp/SKILL.md

@@ -3,42 +3,15 @@ name: speccrew-get-timestamp
 description: Get current timestamp in various formats for file naming and logging. Use when generating dated filenames, report timestamps, or any scenario requiring consistent datetime formatting.
 ---
 
-# Get Timestamp
+# Trigger Scenarios
 
-Get current timestamp in specified format for consistent datetime handling across all skills.
-
-## User
-
-Any Agent or script requiring timestamp generation
-
-## Input
-
-- `format` (optional): Timestamp format, default is `YYYY-MM-DD-HHmmss`
-  - `YYYY-MM-DD-HHmmss`: Full datetime for filenames (e.g., `2026-03-17-132645`)
-  - `YYYY-MM-DD`: Date only (e.g., `2026-03-17`)
-  - `HHmm`: Time only 24h format (e.g., `1326`)
-  - `ISO`: ISO 8601 format (e.g., `2026-03-17T13:26:45+08:00`)
-
-## Output
-
-- Timestamp string in specified format (printed to stdout)
-
-## Supported Formats
-
-| Format | Description | Example |
-|--------|-------------|---------|
-| `YYYY-MM-DD-HHmmss` | Full datetime for filenames | `2026-03-17-132645` |
-| `YYYY-MM-DD` | Date only | `2026-03-17` |
-| `HHmm` | Time only (24h) | `1326` |
-| `ISO` | ISO 8601 format | `2026-03-17T13:26:45+08:00` |
+- Any Agent or script requires timestamp generation
+- User requests formatted timestamp for file naming or logging
+- Workflow step needs consistent datetime formatting
 
 ## AgentFlow Definition
 
 <!-- @agentflow: SKILL.xml -->
 
-> **REQUIRED**: Before executing this workflow, read the XML workflow specification: `speccrew-workspace/docs/rules/agentflow-spec.md`
-
-## Script Location
-
-- **Node.js**: `scripts/get-timestamp.js` (cross-platform, unified implementation)
-
+> **REQUIRED**: Before executing this workflow, read the XML workflow specification: speccrew-workspace/docs/rules/agentflow-spec.md
+> Then read and execute the XML workflow in SKILL.xml block-by-block as the authoritative execution plan.

package/.speccrew/skills/speccrew-knowledge-bizs-api-analyze/SKILL.md

@@ -4,147 +4,12 @@ description: Analyze a single API controller from source code to extract busines
 tools: Read, Write, Edit, Glob, Grep, Bash
 ---
 
-# API Feature Analysis - Single Controller (XML Block Workflow)
-
-> **CRITICAL CONSTRAINT**: DO NOT create temporary scripts, batch files, or workaround code files (`.py`, `.bat`, `.sh`, `.ps1`, etc.) under any circumstances. If execution encounters errors, STOP and report the exact error. Fixes must be applied to the Skill definition or source scripts — not patched at runtime.
-
-Analyze one specific API controller from source code, extract all business features (API endpoints), and generate feature documentation. This skill operates at controller granularity - one worker per controller file.
-
-## Trigger Scenarios
-
+# Trigger Scenarios
 - "Analyze API controller {fileName} from source code"
 - "Extract API features from controller {fileName}"
 - "Generate documentation for API controller {fileName}"
-- "Analyze API feature from features.json"
-
-## Input Variables
-
-| Variable | Type | Description | Example |
-|----------|------|-------------|---------|
-| `${feature}` | object | Complete feature object from features.json | - |
-| `${fileName}` | string | Controller file name | `"UserController"`, `"OrderController"` |
-| `${sourcePath}` | string | Relative path to source file | `"yudao-module-system/yudao-module-system-biz/src/main/java/cn/iocoder/yudao/module/system/controller/admin/user/UserController.java"` |
-| `${documentPath}` | string | Target path for generated document | `"speccrew-workspace/knowledges/bizs/admin-api/system/user/UserController.md"` |
-| `${module}` | string | Business module name (from feature.module) | `"system"`, `"trade"`, `"_root"` |
-| `${analyzed}` | boolean | Analysis status flag | `true` / `false` |
-| `${platform_type}` | string | Platform type | `"admin-api"`, `"app-api"` |
-| `${platform_subtype}` | string | Platform subtype | `"spring-boot"`, `"java"` |
-| `${tech_stack}` | array | Platform tech stack | `["java", "spring-boot", "mybatis-plus"]` |
-| `${completed_dir}` | string | Marker files output directory | `"speccrew-workspace/knowledges/base/sync-state/knowledge-bizs/completed"` |
-| `${sourceFile}` | string | Source features JSON file name | `"features-admin-api.json"` |
-| `${language}` | string | Target language for content | `"zh"`, `"en"` |
-
-## Language Adaptation
-
-**CRITICAL**: Generate all content in the language specified by the `${language}` parameter.
-
-- `${language} == "zh"` → Generate all content in 中文
-- `${language} == "en"` → Generate all content in English
-- Other languages → Use the specified language
-
-**All output content (feature names, descriptions, business rules) must be in the target language only.**
-
-## Output Variables
-
-| Variable | Type | Description |
-|----------|------|-------------|
-| `${status}` | string | Analysis status: `"success"`, `"partial"`, or `"failed"` |
-| `${feature_name}` | string | Name of the analyzed controller |
-| `${generated_file}` | string | Path to the generated documentation file |
-| `${message}` | string | Summary message for status update |
-
-## Execution Requirements
-
-This skill operates in **strict sequential execution mode**:
-- Execute steps in exact order (Step 1 → Step 2 → ... → Step 7)
-- Output step status after each step completion
-- Do NOT skip any step
-
-## Output
-
-**Generated Files:**
-1. `${documentPath}` - Controller documentation file
-2. `${completed_dir}/{module}-{subpath}-{fileName}.done.json` - Completion status marker
-
-**Graph Data Generation:**
-Graph data (nodes, edges) construction is handled by `speccrew-knowledge-bizs-api-graph` Skill.
-After completing API analysis, dispatch will invoke the graph skill to generate `.graph.json` files.
-
-**See Also:**
-- `speccrew-knowledge-bizs-api-graph` - Constructs knowledge graph data from API analysis results
-
-**Return Value:**
-```json
-{
-  "status": "success|partial|failed",
-  "feature": {
-    "fileName": "UserController",
-    "sourcePath": "yudao-module-system/.../controller/admin/user/UserController.java"
-  },
-  "platformType": "admin-api",
-  "module": "system",
-  "featureName": "user-management-api",
-  "generatedFile": "speccrew-workspace/knowledges/bizs/admin-api/system/user/UserController.md",
-  "message": "Successfully analyzed UserController with 8 API endpoints"
-}
-```
-
-The return value is used by dispatch to update the feature status in `features-{platform}.json`.
-
-## Execution Checklist
-
-Before executing the workflow, verify the following inputs:
-
-- Controller: `${fileName}` (`${sourcePath}`)
-- Target: `${documentPath}`
-- Language: `${language}`
-- Module: `${module}`
-- Platform: `${platform_type}`/`${platform_subtype}`
-- Completed Dir: `${completed_dir}`
-- Source File: `${sourceFile}`
 
 ## AgentFlow Definition
-
 <!-- @agentflow: SKILL.xml -->
-
-## Reference Guides
-
-### Mermaid Guide
-
-When generating Mermaid diagrams, follow compatibility guidelines:
-- Use `graph TB` or `graph LR` syntax (not `flowchart`)
-- No parentheses `()` in node text
-- No HTML tags like `<br/>`
-- No `style` definitions
-
-### Business Flow Guidelines
-
-- One diagram per API request
-- Focus on business operations
-- Refer to `templates/FEATURE-DETAIL-TEMPLATE.md`
-
-## Constraints
-
-1. **DO NOT analyze files outside the specified `${sourcePath}`**
-2. **DO NOT generate separate documents for internal/private methods**
-3. **All content MUST be in the language specified by `${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. **Marker files MUST follow naming convention: `{module}-{subpath}-{fileName}.done.json`**
-8. **fileName in .done.json MUST NOT include file extension**
-9. **ALL paths in JSON MUST be relative, not absolute**
-
-## Checklist
-
-- [ ] Template file selected based on `${tech_stack}`
-- [ ] Template content read successfully
-- [ ] Controller file read and analyzed
-- [ ] API endpoints extracted with business flows
-- [ ] API consumers found
-- [ ] Template copied to document path
-- [ ] All sections filled using search_replace
-- [ ] All content in target language (`${language}`)
-- [ ] Results reported in JSON format
-- [ ] .done.json marker file written successfully
-- [ ] .graph.json generation dispatched to graph skill
+> **REQUIRED**: Before executing this workflow, read the XML workflow specification: speccrew-workspace/docs/rules/agentflow-spec.md
+> Then read and execute the XML workflow in SKILL.xml block-by-block as the authoritative execution plan.

package/.speccrew/skills/speccrew-knowledge-bizs-api-graph/SKILL.md

@@ -4,292 +4,12 @@ description: Constructs knowledge graph data (nodes, edges, relationships) from
 tools: Read, Write, Glob, Grep, Bash
 ---
 
-# API Knowledge Graph Constructor (XML Workflow)
-
-> **CRITICAL CONSTRAINT**: DO NOT create temporary scripts, batch files, or workaround code files (`.py`, `.bat`, `.sh`, `.ps1`, etc.) under any circumstances. If execution encounters errors, STOP and report the exact error. Fixes must be applied to the Skill definition or source scripts — not patched at runtime.
-
-Construct knowledge graph data structures (nodes and edges) from API analysis results. This skill transforms structured API documentation into graph JSON format for knowledge base integration.
-
-## Language Adaptation
-
-This skill automatically adapts to the user's input language. All documentation and output will be generated in the same language as the user's query.
-
-## Trigger Scenarios
-
+# Trigger Scenarios
 - "Construct graph data from API analysis results"
 - "Generate knowledge graph nodes and edges for API feature"
 - "Write graph JSON for API controller"
 
-## Input Parameters
-
-| Parameter | Required | Description | Example |
-|-----------|----------|-------------|---------|
-| `api_analysis_path` | Yes | Path to the API analysis document (from bizs-api-analyze) | `"speccrew-workspace/knowledges/bizs/admin-api/system/user/UserController.md"` |
-| `platform_id` | Yes | Target platform identifier | `"admin-api"`, `"app-api"` |
-| `output_dir` | Yes | Output directory for graph data | `"speccrew-workspace/knowledges/base/sync-state/knowledge-bizs/completed"` |
-| `module` | Yes | Business module name | `"system"`, `"trade"`, `"ai"` |
-| `fileName` | Yes | Controller class name (without extension) | `"UserController"` |
-| `sourcePath` | Yes | Relative path to source file | `"yudao-module-system/.../UserController.java"` |
-| `sourceFile` | Yes | Source features JSON filename | `"features-admin-api.json"` |
-| `language` | Yes | Target language for content | `"zh"`, `"en"` |
-| `subpath` | No | Subpath extracted from sourcePath (for marker naming) | `"controller-admin-user"` |
-
-## Output Variables
-
-| Variable | Type | Description |
-|----------|------|-------------|
-| `{{status}}` | string | Graph construction status: `"success"` or `"failed"` |
-| `{{graph_file}}` | string | Path to the generated graph JSON file |
-| `{{node_count}}` | integer | Number of nodes generated |
-| `{{edge_count}}` | integer | Number of edges generated |
-
-## Execution Requirements
-
-This skill operates in **strict sequential execution mode**:
-- Execute steps in exact order (Step 1 → Step 2 → ... → Step 6)
-- Output step status after each step completion
-- Do NOT skip any step
-
-## Output
-
-**Generated Files:**
-1. `{{output_dir}}/{module}-{subpath}-{fileName}.graph.json` - Graph data with nodes and edges
-2. `{{output_dir}}/{module}-{subpath}-{fileName}.graph-done.json` - Graph completion marker
-
-**Return Value:**
-```json
-{
-  "status": "success|failed",
-  "module": "{{module}}",
-  "fileName": "{{fileName}}",
-  "graphFile": "{{output_dir}}/{module}-{subpath}-{fileName}.graph.json",
-  "nodeCount": 15,
-  "edgeCount": 23,
-  "message": "Generated graph data with 15 nodes and 23 edges"
-}
-```
-
 ## AgentFlow Definition
-
 <!-- @agentflow: SKILL.xml -->
-
-## Node ID Naming Convention
-
-```
-{type}-{module}-{name}
-
-Examples:
-  api-system-user-list
-  api-system-user-create
-  service-system-user-service
-  table-system-system_user
-  dto-system-user-create-req
-```
-
-## Marker File Naming Convention
-
-```
-{output_dir}/{module}-{subpath}-{fileName}.graph.json
-```
-
-**How to Extract Each Component:**
-
-1. **module**: Use `{{module}}` input variable directly (e.g., `system`, `trade`, `ai`)
-
-2. **subpath**: Extract from `{{sourcePath}}`:
-   - For Java: Remove package prefix up to the business layer (e.g., `controller/admin/`, `controller/app/`)
-   - Remove the file name at the end
-   - Replace path separators (`/`) with hyphens (`-`)
-   - If the file is at module root, subpath will be empty → omit from filename
-
-3. **fileName**: Use `{{fileName}}` input variable (class name WITHOUT extension)
-
-**Examples:**
-
-| sourcePath | module | subpath | fileName | Marker Filename |
-|------------|--------|---------|----------|-----------------|
-| `yudao-module-system/.../controller/admin/notify/NotifyMessageController.java` | `system` | `controller-admin-notify` | `NotifyMessageController` | `system-controller-admin-notify-NotifyMessageController.graph.json` |
-| `yudao-module-system/.../controller/admin/user/UserController.java` | `system` | `controller-admin-user` | `UserController` | `system-controller-admin-user-UserController.graph.json` |
-| `yudao-module-ai/.../controller/admin/chat/ChatConversationController.java` | `ai` | `controller-admin-chat` | `ChatConversationController` | `ai-controller-admin-chat-ChatConversationController.graph.json` |
-
-## Node Type Reference
-
-| Type | Description | Required Metadata |
-|------|-------------|-------------------|
-| `api` | API endpoint | `method`, `path`, `permissions` |
-| `service` | Service class | `methods` |
-| `table` | Database table | `fields`, `indexes` |
-| `dto` | Data Transfer Object | `fields`, `validation` |
-
-## Edge Type Reference
-
-| Type | Description | Source → Target |
-|------|-------------|-----------------|
-| `operates` | API operates on table | api → table |
-| `invokes` | API calls service | api → service |
-| `references` | API uses DTO | api → dto |
-| `depends-on` | Service dependency | service → service |
-| `maps-to` | DTO maps to table | dto → table |
-
-## Node Structure Examples
-
-### API Node Example
-```json
-{
-  "id": "api-{module}-{endpoint-name}",
-  "type": "api",
-  "name": "<display name>",
-  "module": "{{module}}",
-  "sourcePath": "{{sourcePath}}",
-  "documentPath": "{{api_analysis_path}}",
-  "description": "...",
-  "metadata": {
-    "method": "GET",
-    "path": "/admin-api/system/user/page",
-    "permissions": ["system:user:query"]
-  }
-}
-```
-
-### Service Node Example
-```json
-{
-  "id": "service-{module}-{service-name}",
-  "type": "service",
-  "name": "UserService",
-  "module": "{{module}}",
-  "sourcePath": "relative/path/to/UserService.java",
-  "description": "User business logic service",
-  "metadata": {
-    "methods": ["getUserPage", "createUser", "updateUser"]
-  }
-}
-```
-
-### Table Node Example
-```json
-{
-  "id": "table-{module}-{table-name}",
-  "type": "table",
-  "name": "system_user",
-  "module": "{{module}}",
-  "sourcePath": "",
-  "description": "User table",
-  "metadata": {
-    "fields": ["id", "username", "password", "status"],
-    "indexes": ["idx_username"]
-  }
-}
-```
-
-### DTO Node Example
-```json
-{
-  "id": "dto-{module}-{dto-name}",
-  "type": "dto",
-  "name": "UserCreateReqVO",
-  "module": "{{module}}",
-  "sourcePath": "relative/path/to/UserCreateReqVO.java",
-  "description": "Create user request DTO",
-  "metadata": {
-    "fields": ["username", "password", "nickname"],
-    "validation": ["@NotBlank username", "@Size(max=50) nickname"]
-  }
-}
-```
-
-## Edge Structure Examples
-
-### API to Table Edge (operates)
-```json
-{
-  "source": "api-system-user-list",
-  "target": "table-system-system_user",
-  "type": "operates",
-  "metadata": {
-    "operation": "SELECT",
-    "description": "Query user list with pagination"
-  }
-}
-```
-
-### API to Service Edge (invokes)
-```json
-{
-  "source": "api-system-user-create",
-  "target": "service-system-user-service",
-  "type": "invokes",
-  "metadata": {
-    "method": "createUser",
-    "description": "Create user business logic"
-  }
-}
-```
-
-### API to DTO Edge (references)
-```json
-{
-  "source": "api-system-user-create",
-  "target": "dto-system-user-create-req",
-  "type": "references",
-  "metadata": {
-    "usage": "request",
-    "description": "Create user request body"
-  }
-}
-```
-
-### Service to Service Edge (depends-on)
-```json
-{
-  "source": "service-system-user-service",
-  "target": "service-system-permission-service",
-  "type": "depends-on",
-  "metadata": {
-    "description": "User service depends on permission service for role checks"
-  }
-}
-```
-
-### DTO to Table Edge (maps-to)
-```json
-{
-  "source": "dto-system-user-do",
-  "target": "table-system-system_user",
-  "type": "maps-to",
-  "metadata": {
-    "description": "UserDO maps to system_user table"
-  }
-}
-```
-
-## Pre-write Verification Checklist
-
-### Graph JSON Verification:
-- [ ] Root-level `module` field is present (MANDATORY)
-- [ ] `nodes` and `edges` are arrays (can be empty)
-- [ ] Valid JSON (no trailing commas, all strings quoted)
-- [ ] All node IDs are unique
-- [ ] All edge source/target references point to valid node IDs
-
-### API Endpoint Coverage Check:
-- [ ] ALL public API endpoint methods in the controller are represented as `api` nodes
-- [ ] Status update endpoints (updateStatus, toggleEnable) are included
-- [ ] Special operation endpoints (resetPassword, export, import, batch operations) are included
-- [ ] Each `api` node has proper metadata with HTTP method and path
-- [ ] No public endpoint method is left without a corresponding node
-
-### Completion Marker Verification:
-- [ ] Filename follows `{module}-{subpath}-{fileName}.graph-done.json` pattern
-- [ ] JSON is valid
-- [ ] All required fields are present
-- [ ] `nodeCount` and `edgeCount` match actual graph data
-
-## Constraints
-
-1. **Single Document Input**: This skill processes ONE API analysis document at a time
-2. **JSON Format**: All output files MUST be valid JSON
-3. **Module Field**: The root-level `module` field is MANDATORY in graph JSON
-4. **Node Uniqueness**: Each node ID must be unique within the graph
-5. **Edge Validity**: Edge source/target must reference existing node IDs
-6. **Path Format**: Use relative paths, NEVER absolute paths in JSON content
+> **REQUIRED**: Before executing this workflow, read the XML workflow specification: speccrew-workspace/docs/rules/agentflow-spec.md
+> Then read and execute the XML workflow in SKILL.xml block-by-block as the authoritative execution plan.