refactor(workflow): enhance init command documentation with memory discovery features

2026-02-05 01:50:27 +08:00 · 2025-11-19 20:46:37 +08:00
parent 487b359266
commit 9b07310d68
1 changed files with 173 additions and 8 deletions
--- a/.claude/commands/workflow/init.md
+++ b/.claude/commands/workflow/init.md
@@ -10,12 +10,14 @@ examples:
 # Workflow Init Command (/workflow:init)

 ## Overview
-Initializes `.workflow/project.json` with comprehensive project understanding by leveraging **cli-explore-agent** for intelligent analysis.
+Initializes `.workflow/project.json` with comprehensive project understanding by leveraging **cli-explore-agent** for intelligent analysis and **memory discovery** for SKILL package indexing.

 **Key Features**:
 - **Intelligent Project Analysis**: Uses cli-explore-agent's Deep Scan mode
 - **Technology Stack Detection**: Identifies languages, frameworks, build tools
 - **Architecture Overview**: Discovers patterns, layers, key components
+- **Memory Discovery**: Scans and indexes available SKILL packages
+- **Smart Recommendations**: Suggests memory commands based on project state
 - **One-time Initialization**: Skips if project.json exists (unless --regenerate)

 ## Usage
@@ -246,13 +248,19 @@ Files: ${metrics.total_files}
 LOC: ${metrics.lines_of_code}
 Complexity: ${metrics.complexity}

-## Next Steps
-1. Start a workflow: /workflow:plan "feature description"
-2. View project state: /workflow:status --project
-3. View details: cat .workflow/project.json
+### Memory Resources
+SKILL Packages: ${memory_resources.skills.length}
+Documentation: ${memory_resources.documentation.length} project(s)
+Module Docs: ${memory_resources.module_docs.length} file(s)
+Gaps: ${memory_resources.gaps.join(', ') || 'none'}
+
+## Quick Start
+• /workflow:plan "feature description" - Start new workflow
+• /workflow:status --project - View project state

 ---
 Project state saved to: .workflow/project.json
+Memory index updated: ${memory_resources.last_scanned}
 ```

 ## Extended project.json Schema
@@ -321,14 +329,130 @@ Project state saved to: .workflow/project.json
    "last_updated": "2025-01-18T10:00:00Z"
  },

+  "memory_resources": {
+    "skills": [
+      {"name": "claude_dms3", "type": "project_docs", "path": ".claude/skills/claude_dms3"},
+      {"name": "workflow-progress", "type": "workflow_progress", "path": ".claude/skills/workflow-progress"}
+    ],
+    "documentation": [
+      {
+        "name": "claude_dms3",
+        "path": ".workflow/docs/claude_dms3",
+        "has_readme": true,
+        "has_architecture": true
+      }
+    ],
+    "module_docs": [
+      ".claude/commands/workflow/CLAUDE.md",
+      ".claude/agents/CLAUDE.md"
+    ],
+    "gaps": ["tech_stack"],
+    "last_scanned": "2025-01-18T10:05:00Z"
+  },
+
  "_metadata": {
    "initialized_by": "cli-explore-agent",
    "analysis_timestamp": "2025-01-18T10:00:00Z",
-    "analysis_mode": "deep-scan"
+    "analysis_mode": "deep-scan",
+    "memory_scan_timestamp": "2025-01-18T10:05:00Z"
  }
 }
 ```

+### Phase 5: Discover Memory Resources
+
+**Goal**: Scan and index available SKILL packages (memory command products) using agent delegation
+
+**Invoke general-purpose agent** to discover and catalog all memory products:
+
+```javascript
+Task(
+  subagent_type="general-purpose",
+  description="Discover memory resources",
+  prompt=`
+Discover and index all memory command products: SKILL packages, documentation, and CLAUDE.md files.
+
+## Discovery Scope
+1. **SKILL Packages** (.claude/skills/) - Generated by /memory:skill-memory, /memory:tech-research, etc.
+2. **Documentation** (.workflow/docs/) - Generated by /memory:docs
+3. **Module Docs** (**/CLAUDE.md) - Generated by /memory:update-full, /memory:update-related
+
+## Discovery Tasks
+
+### 1. Scan SKILL Packages
+- List all directories in .claude/skills/
+- For each: extract name, classify type, record path
+- Types: workflow-progress | codemap-* | style-* | tech_stacks | project_docs
+
+### 2. Scan Documentation
+- List directories in .workflow/docs/
+- For each project: name, path, check README.md, ARCHITECTURE.md existence
+
+### 3. Scan CLAUDE.md Files
+- Find all **/CLAUDE.md (exclude: node_modules, .git, dist, build)
+- Return path list only
+
+### 4. Identify Gaps
+- No project SKILL? → "project_skill"
+- No documentation? → "documentation"
+- Missing tech stack SKILL? → "tech_stack"
+- No workflow-progress? → "workflow_history"
+- <10% modules have CLAUDE.md? → "module_docs_low_coverage"
+
+### 5. Return JSON:
+{
+  "skills": [
+    {"name": "claude_dms3", "type": "project_docs", "path": ".claude/skills/claude_dms3"},
+    {"name": "workflow-progress", "type": "workflow_progress", "path": ".claude/skills/workflow-progress"}
+  ],
+  "documentation": [
+    {
+      "name": "my_project",
+      "path": ".workflow/docs/my_project",
+      "has_readme": true,
+      "has_architecture": true
+    }
+  ],
+  "module_docs": [
+    "src/core/CLAUDE.md",
+    "lib/utils/CLAUDE.md"
+  ],
+  "gaps": ["tech_stack", "module_docs_low_coverage"]
+}
+
+## Context
+- Project tech stack: ${JSON.stringify(analysis.technology_stack)}
+- Check .workflow/.archives for session history
+- If directories missing, return empty state with recommendations
+  `
+)
+```
+
+**Agent Output**: JSON structure with skills, documentation, module_docs, and gaps
+
+**Update project.json**:
+```javascript
+const memoryDiscovery = JSON.parse(agentOutput);
+
+projectMeta.memory_resources = {
+  ...memoryDiscovery,
+  last_scanned: new Date().toISOString()
+};
+
+Write('.workflow/project.json', JSON.stringify(projectMeta, null, 2));
+```
+
+**Output Summary**:
+```
+Memory Resources Indexed:
+- SKILL Packages: ${skills.length}
+- Documentation: ${documentation.length} project(s)
+- Module Docs: ${module_docs.length} file(s)
+- Gaps: ${gaps.join(', ') || 'none'}
+```
+
+---
+
 ## Regeneration Behavior

 When using `--regenerate` flag:
@@ -347,7 +471,9 @@ When using `--regenerate` flag:

 3. **Re-run cli-explore-agent analysis**

-4. **Merge preserved data with new analysis**:
+4. **Re-run memory discovery (Phase 5)**
+
+5. **Merge preserved data with new analysis**:
   ```javascript
   const newProjectMeta = {
     ...analysisResults,
@@ -356,7 +482,7 @@ When using `--regenerate` flag:
   };
   ```

-5. **Output**:
+6. **Output**:
   ```
   ✓ Project analysis regenerated
   Backup saved: .workflow/project.json.backup
@@ -365,6 +491,7 @@ When using `--regenerate` flag:
   - Technology stack analysis
   - Architecture overview
   - Key components discovery
+   - Memory resources index

   Preserved:
   - ${preservedFeatures.length} existing features
@@ -397,3 +524,41 @@ If not in git repo and empty directory:
 2. Create minimal project.json
 3. Suggest: "Add code files and run /workflow:init --regenerate"
 ```
+
+### Memory Discovery Failures
+
+**Missing Directories**:
+```
+If .claude/skills, .workflow/docs, or CLAUDE.md files not found:
+1. Return empty state for that category
+2. Mark in gaps.missing array
+3. Continue initialization
+```
+
+**Metadata Read Failures**:
+```
+If SKILL.md files are unreadable:
+1. Include SKILL with basic info: name (from directory), type (inferred), path
+2. Log warning: "SKILL package {name} has invalid metadata"
+3. Continue with other SKILLs
+```
+
+**Coverage Check Failures**:
+```
+If unable to determine module doc coverage:
+1. Skip adding "module_docs_low_coverage" to gaps
+2. Continue with other gap checks
+```
+
+**Default Empty State**:
+```json
+{
+  "memory_resources": {
+    "skills": [],
+    "documentation": [],
+    "module_docs": [],
+    "gaps": ["project_skill", "documentation", "tech_stack", "workflow_history", "module_docs"],
+    "last_scanned": "ISO_TIMESTAMP"
+  }
+}
+```