[IMPORTANT] Use TaskCreate to break ALL work into small tasks BEFORE starting — including tasks for each file read. This prevents context loss from long files. For simple tasks, AI MUST ATTENTION ask user whether to skip.Prerequisites: MUST ATTENTION READ before executing:
Scan & Update Reference Doc — When updating reference docs: (1) Read existing doc first. (2) Scan codebase for current state (grep/glob). (3) Diff findings vs doc content. (4) Update ONLY sections where code diverged from doc. (5) Preserve manual annotations. (6) Update metadata (date, counts). NEVER rewrite entire doc — surgical updates only.
Output Quality — 10 rules for reference docs: (1) No inventories/counts, (2) No directory trees, (3) No TOCs, (4) Rules over descriptions, (5) 1 example per pattern, (6) Tables over prose, (7) Primacy-recency anchoring (critical rules in first+last 5 lines), (8) No checkbox checklists — use "MUST ATTENTION verify X", (9) Min density: 8 MUST ATTENTION/NEVER/ALWAYS per 100 lines, (10) Verify base class names and code examples preserved.
Quick Summary
Goal: Scan existing business feature documentation and populate docs/project-reference/feature-docs-reference.md with app-to-service mapping, documentation structure conventions, template usage, and section standards.
Workflow:
- Read — Load current target doc, detect init vs sync mode
- Scan — Discover documentation patterns via parallel sub-agents
- Report — Write findings to external report file
- Generate — Build/update reference doc from report
- Verify — Validate discovered paths and templates exist
Key Rules:
- Generic — works with any documentation structure (docs/, wiki/, etc.)
- Discover documentation organization dynamically from file system
- Map relationships between apps/services and their documentation
- Every reference must point to real files found in this project
Be skeptical. Apply critical thinking, sequential thinking. Every claim needs traced proof, confidence percentages (Idea should be more than 80%).
Scan Feature Docs
Phase 0: Read & Assess
- Read
docs/project-reference/feature-docs-reference.md - Detect mode: init (placeholder) or sync (populated)
- If sync: extract existing sections and note what's already well-documented
Phase 1: Plan Scan Strategy
Discover documentation locations:
docs/directory structure (business features, architecture, guides)docs/business-features/or similar feature doc directoriesdocs/templates/or similar template directories- README.md files across service directories
- Wiki or external doc references in config files
Use docs/project-config.json if available for module lists and app mappings.
Phase 2: Execute Scan (Parallel Sub-Agents)
Launch 2 Explore agents in parallel:
Agent 1: Documentation Structure
- Glob for
docs/**/*.mdto map full documentation tree - Find documentation templates (template files, skeleton docs)
- Discover documentation section patterns (recurring H2/H3 headings across docs)
- Identify INDEX.md / README.md hub files and their link structures
- Count docs per app/module to assess coverage distribution
- Identify documentation naming patterns across feature docs
Agent 2: App-to-Service Mapping
- Map frontend apps to backend services (from config, imports, or API calls)
- Find API reference docs and their relationship to services
- Discover troubleshooting docs and their coverage
- Find cross-references between docs (links, mentions)
- Identify documentation conventions (naming, numbering, tagging patterns)
- Look for doc generation tools or scripts
Write all findings to: plans/reports/scan-feature-docs-{YYMMDD}-{HHMM}-report.md
Phase 3: Analyze & Generate
Read the report. Build these sections:
Target Sections
| Section | Content |
|---|---|
| App-to-Service Mapping | Table: App name, Backend services, Doc directory, Doc count |
| Directory Structure | Tree showing docs/ organization with purpose annotations |
| Template Paths | Table: Template name, Path, Purpose, Used by N docs |
| Section Structure | Standard sections found across feature docs (with frequency) |
| Documentation Conventions | Naming conventions, numbering schemes, required fields |
| Evidence Rules | How docs reference code (file:line patterns, test case IDs) |
| Coverage Gaps | Apps/services without documentation, incomplete docs |
Content Rules
- Use tables for all structured data (mappings, templates, conventions)
- Include actual directory tree output (top 3 levels)
- Show section heading patterns with frequency counts
- Highlight well-documented vs under-documented areas
Phase 4: Write & Verify
- Write updated doc with
<!-- Last scanned: YYYY-MM-DD -->at top - Verify: 3 template paths exist on filesystem
- Verify: app-to-service mappings match actual directory structure
- Report: sections updated, coverage statistics, gaps identified
Closing Reminders
- IMPORTANT MUST ATTENTION break work into small todo tasks using
TaskCreateBEFORE starting - IMPORTANT MUST ATTENTION search codebase for 3+ similar patterns before creating new code
- IMPORTANT MUST ATTENTION cite
file:lineevidence for every claim (confidence >80% to act) - IMPORTANT MUST ATTENTION add a final review todo task to verify work quality MANDATORY IMPORTANT MUST ATTENTION READ the following before starting:
- IMPORTANT MUST ATTENTION read existing doc first, scan codebase, diff, surgical update only. Never rewrite entire doc.