@su-record/vibe 2.6.14 → 2.6.15

package/README.md

@@ -1,497 +1,497 @@
-# Vibe
-
-**SPEC-Scenario-driven AI coding framework for Claude Code with Multi-LLM orchestration**  
-
-[![npm version](https://img.shields.io/npm/v/@su-record/vibe.svg)](https://www.npmjs.com/package/@su-record/vibe)
-[![npm downloads](https://img.shields.io/npm/dt/@su-record/vibe)](https://www.npmjs.com/package/@su-record/vibe)
-[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
-
-## Install
-
-```bash
-npm install -g @su-record/vibe
-vibe init
-```
-
-## Commands
-
-### Terminal
-
-| Command | Description |
-|---------|-------------|
-| `vibe init` | Initialize project |
-| `vibe update` | Update settings |
-| `vibe status` | Check status |
-| `vibe gpt auth` | GPT OAuth authentication |
-| `vibe gpt key <KEY>` | GPT API key setup |
-| `vibe gpt logout` | GPT logout |
-| `vibe gemini auth` | Gemini OAuth authentication |
-| `vibe gemini key <KEY>` | Gemini API key setup |
-| `vibe gemini logout` | Gemini logout |
-| `vibe help` | Help |
-| `vibe version` | Version info |
-
-### Claude Code
-
-| Command | Description |
-|---------|-------------|
-| `/vibe.spec "feature"` | Create SPEC + parallel research |
-| `/vibe.run "feature"` | Execute implementation |
-| `/vibe.run "feature" ultrawork` | Maximum performance mode |
-| `/vibe.verify "feature"` | BDD verification |
-| `/vibe.trace "feature"` | Requirements traceability matrix (v2.6) |
-| `/vibe.review` | 13+ agent parallel review (`--race` for GPT+Gemini) |
-| `/vibe.analyze` | Code analysis |
-| `/vibe.reason "problem"` | Systematic reasoning |
-| `/vibe.utils --ui` | UI preview (Gemini image / ASCII fallback) |
-| `/vibe.utils --e2e` | E2E testing, diagrams, etc. |
-
-## Workflow
-
-```
-/vibe.spec → /vibe.run → /vibe.trace → /vibe.verify → /vibe.review
-                              ↑
-                     Coverage check (v2.6)
-```
-
-## Key Features
-
-| Feature | Description |
-|---------|-------------|
-| **Race Review (v2.6.9)** | GPT + Gemini parallel review with cross-validation |
-| **Fire-and-Forget Agents (v2.6)** | Non-blocking background execution with instant handle return |
-| **Phase Pipelining (v2.6)** | Prepare next phase during current execution |
-| **PRD-to-SPEC Automation (v2.6)** | Auto-generate SPEC from PRD documents |
-| **Traceability Matrix (v2.6)** | REQ → SPEC → Feature → Test coverage tracking |
-| **Multi-LLM Research (v2.5)** | Claude + GPT + Gemini 3-way validation during SPEC research |
-| **Multi-model orchestration** | Claude + GPT-5.2 + Gemini 3 Pro |
-| **13+ parallel review agents** | Security, performance, architecture |
-| **Iterative SPEC review** | GPT + Gemini convergence loop (max 3 rounds) |
-| **Large SPEC auto-split** | 5+ phases auto-split into phase files |
-| **BDD auto verification** | Given/When/Then scenario verification |
-| **ULTRAWORK mode** | One keyword enables all optimizations |
-| **30+ built-in tools** | Code analysis, memory management, quality validation |
-| **Auto context management** | 80%+ auto-save, session restore |
-| **23 language presets** | TypeScript, Python, Go, Rust, Swift, Kotlin, C#, Ruby, Dart, GDScript |
-
-## ULTRAWORK Mode
-
-Enable maximum performance with `ultrawork` or `ulw`:
-
-```bash
-/vibe.run "feature" ultrawork
-/vibe.run "feature" ulw        # Same
-```
-
-**Enabled features:**
-- Parallel subagent exploration (3+ concurrent)
-- **Background agents** (v2.6) - Prepare next phase during implementation
-- **Phase pipelining** (v2.6) - Zero wait time between phases
-- Boulder Loop (auto-progress until all scenarios complete)
-- Auto-retry on errors (max 3)
-- Continuous execution without phase confirmation
-- Auto-save at 80%+ context
-
-## Fire-and-Forget Background Manager (v2.6)
-
-Non-blocking task execution with instant response:
-
-```typescript
-import { launch, poll, getStats } from '@su-record/vibe/orchestrator';
-
-// Fire-and-forget: returns immediately (<100ms)
-const { taskId } = launch({
-  prompt: 'Analyze codebase security',
-  agentName: 'security-agent',
-  model: 'claude-sonnet-4-5'
-});
-
-// Poll for result later
-const result = await poll(taskId);
-
-// Check queue stats
-const stats = getStats();
-```
-
-**Concurrency limits:**
-
-| Model  | Concurrent Limit |
-|--------|------------------|
-| Opus   | 3                |
-| Sonnet | 5                |
-| Haiku  | 8                |
-
-**Queue features:**
-
-- Bounded queue (max 100 tasks)
-- Per-model rate limiting
-- Task timeout: 180s individual, 10min pipeline
-- Automatic cleanup (24h TTL)
-
-## Phase Pipelining (v2.6)
-
-Prepare next phase while current phase executes:
-
-```
-Phase N execution     │ Background: Phase N+1 preparation
-──────────────────────┼──────────────────────────────────
-[Implementing...]     │ [Analyzing files for next phase]
-[Testing...]          │ [Pre-generating test cases]
-[Complete]            │ [Ready to start immediately!]
-                      ↓
-Phase N+1 starts with ZERO wait time
-```
-
-**Speed comparison:**
-
-| Mode                      | 5 Phases Total |
-|---------------------------|----------------|
-| Sequential                | ~10min         |
-| Parallel Exploration      | ~7.5min        |
-| **ULTRAWORK + Pipeline**  | **~5min**      |
-
-## PRD-to-SPEC Automation (v2.6)
-
-Auto-generate SPEC from PRD documents:
-
-```typescript
-import { parsePRD } from '@su-record/vibe/tools';
-
-// Parse PRD (supports Markdown, YAML frontmatter)
-const prd = parsePRD(prdContent, 'login-feature');
-
-// Extracted requirements with auto-generated IDs
-// REQ-login-001, REQ-login-002, ...
-```
-
-**Requirement ID System:**
-
-- Format: `REQ-{feature}-{number}`
-- Auto-deduplication
-- Priority inference from keywords (must, should, nice-to-have)
-
-## Requirements Traceability Matrix (v2.6)
-
-Track coverage from requirements to tests:
-
-```bash
-/vibe.trace "feature-name"  # Generate RTM
-```
-
-**RTM tracks:**
-
-```
-REQ-login-001 → SPEC Phase 1 → Feature Scenario 1 → login.test.ts
-REQ-login-002 → SPEC Phase 2 → Feature Scenario 3 → auth.test.ts
-```
-
-**Coverage report:**
-
-- Requirements coverage %
-- Missing implementations
-- Untested features
-- Export: Markdown, HTML
-
-## SPEC Versioning (v2.6)
-
-Automatic version management for SPEC documents:
-
-```typescript
-import { bumpVersion, generateChangelog } from '@su-record/vibe/tools';
-
-// Auto-bump version on SPEC changes
-await bumpVersion('feature-name', 'minor');
-
-// Generate changelog
-await generateChangelog('feature-name');
-```
-
-**Features:**
-
-- Semantic versioning (major.minor.patch)
-- Git tag integration
-- Automatic changelog generation
-- Baseline tagging for releases
-
-## UI Preview (`--ui`)
-
-Generate UI previews from text description or design folder:
-
-```bash
-/vibe.utils --ui "login form with email/password"
-/vibe.utils --ui ./design/dashboard/
-```
-
-**Features:**
-
-- **Gemini enabled**: Generates actual UI mockup image
-- **Gemini disabled**: ASCII art fallback
-- Supports: HTML, PNG, JPG, CSS, JSON, SVG, MD
-
-## Multi-model Orchestration
-
-**Hook-based automatic routing** - Keywords in your prompt trigger the right LLM automatically:
-
-| Keyword Pattern | Routes to | Use Case |
-|-----------------|-----------|----------|
-| `architecture`, `design` | GPT | Architecture review |
-| `UI`, `UX` | Gemini | UI/UX feedback |
-| `debugging`, `find bug` | GPT | Bug analysis |
-| `analyze code` | Gemini | Code review |
-
-**Smart Routing features:**
-- Automatic LLM selection based on task type
-- Exponential backoff retry (3 attempts, 2s → 4s → 8s)
-- Auto fallback chain: GPT ↔ Gemini (if primary fails)
-- 5-minute availability cache
-- Web search via Gemini (Google Search grounding)
-
-**API usage:**
-
-```javascript
-import('@su-record/vibe/lib/gpt').then(g => g.ask('question'))
-import('@su-record/vibe/lib/gemini').then(g => g.webSearch('search query'))
-```
-
-## Multi-LLM Research (v2.5)
-
-During `/vibe.spec`, research agents use **3 LLM perspectives** for quality assurance:
-
-```text
-┌─────────────────────────────────────────────────────────────┐
-│  MULTI-LLM PARALLEL RESEARCH                                │
-├─────────────────────────────────────────────────────────────┤
-│  best-practices-agent                                       │
-│    ├── Claude: Core patterns, anti-patterns                 │
-│    ├── GPT: Architecture patterns, code conventions         │
-│    └── Gemini: Latest trends, framework updates             │
-├─────────────────────────────────────────────────────────────┤
-│  security-advisory-agent                                    │
-│    ├── Claude: OWASP Top 10, security patterns              │
-│    ├── GPT: CVE database, vulnerability details             │
-│    └── Gemini: Latest security advisories, patches          │
-└─────────────────────────────────────────────────────────────┘
-```
-
-> vibe = Quality Assurance Framework
-
-## Race Review (v2.6.9)
-
-**Multi-LLM competitive review** - Same task runs on GPT + Gemini in parallel, results cross-validated:
-
-```bash
-/vibe.review --race              # All review types
-/vibe.review --race security     # Security only
-```
-
-**Cross-validation confidence:**
-
-| Agreement         | Priority | Action                  |
-|-------------------|----------|-------------------------|
-| Both agree (100%) | P1       | High confidence finding |
-| One model (50%)   | P2       | Needs verification      |
-
-**Output:**
-
-```text
-🏁 RACE REVIEW
-├─ GPT:    [SQL injection, XSS]
-└─ Gemini: [SQL injection, CSRF]
-         ↓
-Cross-validation:
-- SQL injection (2/2) → 🔴 P1 (100%)
-- XSS (1/2) → 🟡 P2 (50%)
-- CSRF (1/2) → 🟡 P2 (50%)
-```
-
-**ULTRAWORK default:** Race review is automatically enabled in ULTRAWORK mode.
-
-**API usage:**
-
-```javascript
-import('@su-record/vibe/tools').then(t =>
-  t.raceReview({ reviewType: 'security', code: 'CODE' })
-   .then(r => console.log(t.formatRaceResult(r)))
-)
-```
-
-## Parallel Review Agents
-
-`/vibe.review` runs 13+ agents concurrently:
-
-| Category | Agents |
-|----------|--------|
-| Security | security-reviewer, data-integrity-reviewer |
-| Performance | performance-reviewer, complexity-reviewer |
-| Architecture | architecture-reviewer, simplicity-reviewer |
-| Language | python, typescript, rails, react reviewers |
-| Context | git-history, test-coverage reviewers |
-
-**Priority:**
-- 🔴 P1 (Critical): Blocks merge
-- 🟡 P2 (Important): Recommended fix
-- 🔵 P3 (Nice-to-have): Backlog
-
-## Built-in Tools (25)
-
-| Category | Tools |
-|----------|-------|
-| **Semantic** | `find_symbol`, `find_references`, `analyze_dependency_graph` |
-| **Convention** | `analyze_complexity`, `validate_code_quality`, `check_coupling_cohesion`, `suggest_improvements`, `apply_quality_rules` |
-| **Memory** | `start_session`, `save_memory`, `auto_save_context`, `recall_memory`, `search_memories`, `search_memories_advanced`, `list_memories`, `update_memory`, `delete_memory`, `link_memories`, `prioritize_memory`, `get_memory_graph`, `create_memory_timeline`, `get_session_context`, `restore_session_context` |
-| **UI** | `preview_ui_ascii` |
-| **Time** | `get_current_time` |
-
-## Project Structure
-
-**Global install (`~/.claude/`):**
-
-```text
-~/.claude/
-├── commands/       # Slash commands (7)
-├── agents/         # Review/research agents
-├── skills/         # Auto-activated guides (7)
-└── settings.json   # Hooks settings
-```
-
-**Project-specific (`project/.claude/vibe/`):**
-
-```text
-.claude/vibe/
-├── specs/          # SPEC documents
-├── features/       # BDD scenarios
-├── rules/          # Coding rules (per stack)
-├── solutions/      # Solution archive
-├── config.json     # Project settings
-└── constitution.md # Project principles
-```
-
-## Cursor IDE Support (v2.6.8)
-
-VIBE automatically installs Cursor-compatible assets on `vibe init` and `vibe update`:
-
-**Installed locations:**
-
-| Asset | Path | Count |
-| ----- | ---- | ----- |
-| Subagents | `~/.cursor/agents/*.md` | 12 |
-| Skills | `~/.cursor/skills/*/SKILL.md` | 7 |
-| Rules Template | `~/.cursor/rules-template/*.mdc` | 5 |
-
-**12 Review Subagents with optimized model mapping:**
-
-| Agent Type | Model | Purpose |
-| ---------- | ----- | ------- |
-| security, architecture, data-integrity | `claude-4.5-sonnet-thinking` | Deep reasoning |
-| typescript, python, react, rails | `gpt-5.2-codex` | Code understanding |
-| performance, complexity, simplicity, test-coverage, git-history | `gemini-3-flash` | Fast pattern checks |
-
-**7 Skills (VIBE workflow guides):**
-
-- `vibe-spec` - SPEC creation
-- `vibe-run` - Implementation execution
-- `vibe-review` - Parallel code review
-- `vibe-analyze` - Project analysis
-- `vibe-verify` - Requirement verification
-- `vibe-reason` - Systematic reasoning
-- `vibe-ui` - UI preview utilities
-
-**5 Rules Templates:**
-
-- `typescript-standards.mdc` - Type safety, complexity limits
-- `react-patterns.mdc` - Component patterns
-- `code-quality.mdc` - General quality rules
-- `security-checklist.mdc` - OWASP checklist
-- `python-standards.mdc` - PEP8, type hints
-
-**Usage in Cursor:**
-
-```text
-# Invoke subagent
-"Use security-reviewer to check this code"
-
-# Use skill workflow
-"vibe spec login feature"
-```
-
-## Code Quality Standards
-
-| Metric | Limit |
-|--------|-------|
-| Function length | 30 lines recommended, 50 allowed |
-| Nesting depth | 3 levels max |
-| Parameters | 5 max |
-| Cyclomatic complexity | 10 max |
-
-## Usage Example
-
-```
-User: /vibe.spec "login feature"
-
-Claude: You're building a login feature! I have a few questions.
-        1. Auth method? (email/password, OAuth, Passkey)
-        2. Tech stack?
-        ...
-
-[Requirements confirmed via conversation]
-[4 parallel research agents run]
-
-✅ SPEC document created
-📄 .claude/vibe/specs/login.md
-📄 .claude/vibe/features/login.feature
-
-Next: /vibe.run "login feature" ultrawork
-```
-
-## Context7 Plugin (Recommended)
-
-Install context7 plugin for up-to-date library documentation:
-
-```bash
-/plugin install context7
-```
-
-**Benefits:**
-
-- **Skill integration**: Auto-invokes context7 for library/API questions without prompting
-- **Subagent isolation**: Docs queries run in separate context, preventing main context bloat
-- **Knowledge cutoff solution**: Access latest docs for long coding sessions
-
-**Usage:**
-
-- Automatic: Ask library questions, vibe's skill triggers context7
-- Manual: `/context7:docs <library> [query]`
-
-## Requirements
-
-- Node.js 18.0.0+
-- Claude Code
-
-## License
-
-MIT - [GitHub](https://github.com/su-record/vibe)
-
-## Vibe Setup (AI Coding)
-
-This project uses [Vibe](https://github.com/su-record/vibe) AI coding framework.
-
-### Collaborator Install
-
-```bash
-# Global install (recommended)
-npm install -g @su-record/vibe
-vibe update
-
-# Or use vibe init to setup
-vibe init
-```
-
-### Usage
-
-Use slash commands in Claude Code:
-- `/vibe.spec "feature"` - Create SPEC document
-- `/vibe.run "feature"` - Execute implementation
-
----
-
-⭐ If this helps your workflow, consider giving it a star!
+# Vibe
+
+**SPEC-Scenario-driven AI coding framework for Claude Code with Multi-LLM orchestration**  
+
+[![npm version](https://img.shields.io/npm/v/@su-record/vibe.svg)](https://www.npmjs.com/package/@su-record/vibe)
+[![npm downloads](https://img.shields.io/npm/dt/@su-record/vibe)](https://www.npmjs.com/package/@su-record/vibe)
+[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
+
+## Install
+
+```bash
+npm install -g @su-record/vibe
+vibe init
+```
+
+## Commands
+
+### Terminal
+
+| Command | Description |
+|---------|-------------|
+| `vibe init` | Initialize project |
+| `vibe update` | Update settings |
+| `vibe status` | Check status |
+| `vibe gpt auth` | GPT OAuth authentication |
+| `vibe gpt key <KEY>` | GPT API key setup |
+| `vibe gpt logout` | GPT logout |
+| `vibe gemini auth` | Gemini OAuth authentication |
+| `vibe gemini key <KEY>` | Gemini API key setup |
+| `vibe gemini logout` | Gemini logout |
+| `vibe help` | Help |
+| `vibe version` | Version info |
+
+### Claude Code
+
+| Command | Description |
+|---------|-------------|
+| `/vibe.spec "feature"` | Create SPEC + parallel research |
+| `/vibe.run "feature"` | Execute implementation |
+| `/vibe.run "feature" ultrawork` | Maximum performance mode |
+| `/vibe.verify "feature"` | BDD verification |
+| `/vibe.trace "feature"` | Requirements traceability matrix (v2.6) |
+| `/vibe.review` | 13+ agent parallel review (`--race` for GPT+Gemini) |
+| `/vibe.analyze` | Code analysis |
+| `/vibe.reason "problem"` | Systematic reasoning |
+| `/vibe.utils --ui` | UI preview (Gemini image / ASCII fallback) |
+| `/vibe.utils --e2e` | E2E testing, diagrams, etc. |
+
+## Workflow
+
+```
+/vibe.spec → /vibe.run → /vibe.trace → /vibe.verify → /vibe.review
+                              ↑
+                     Coverage check (v2.6)
+```
+
+## Key Features
+
+| Feature | Description |
+|---------|-------------|
+| **Race Review (v2.6.9)** | GPT + Gemini parallel review with cross-validation |
+| **Fire-and-Forget Agents (v2.6)** | Non-blocking background execution with instant handle return |
+| **Phase Pipelining (v2.6)** | Prepare next phase during current execution |
+| **PRD-to-SPEC Automation (v2.6)** | Auto-generate SPEC from PRD documents |
+| **Traceability Matrix (v2.6)** | REQ → SPEC → Feature → Test coverage tracking |
+| **Multi-LLM Research (v2.5)** | Claude + GPT + Gemini 3-way validation during SPEC research |
+| **Multi-model orchestration** | Claude + GPT-5.2 + Gemini 3 Pro |
+| **13+ parallel review agents** | Security, performance, architecture |
+| **Iterative SPEC review** | GPT + Gemini convergence loop (max 3 rounds) |
+| **Large SPEC auto-split** | 5+ phases auto-split into phase files |
+| **BDD auto verification** | Given/When/Then scenario verification |
+| **ULTRAWORK mode** | One keyword enables all optimizations |
+| **30+ built-in tools** | Code analysis, memory management, quality validation |
+| **Auto context management** | 80%+ auto-save, session restore |
+| **23 language presets** | TypeScript, Python, Go, Rust, Swift, Kotlin, C#, Ruby, Dart, GDScript |
+
+## ULTRAWORK Mode
+
+Enable maximum performance with `ultrawork` or `ulw`:
+
+```bash
+/vibe.run "feature" ultrawork
+/vibe.run "feature" ulw        # Same
+```
+
+**Enabled features:**
+- Parallel subagent exploration (3+ concurrent)
+- **Background agents** (v2.6) - Prepare next phase during implementation
+- **Phase pipelining** (v2.6) - Zero wait time between phases
+- Boulder Loop (auto-progress until all scenarios complete)
+- Auto-retry on errors (max 3)
+- Continuous execution without phase confirmation
+- Auto-save at 80%+ context
+
+## Fire-and-Forget Background Manager (v2.6)
+
+Non-blocking task execution with instant response:
+
+```typescript
+import { launch, poll, getStats } from '@su-record/vibe/orchestrator';
+
+// Fire-and-forget: returns immediately (<100ms)
+const { taskId } = launch({
+  prompt: 'Analyze codebase security',
+  agentName: 'security-agent',
+  model: 'claude-sonnet-4-5'
+});
+
+// Poll for result later
+const result = await poll(taskId);
+
+// Check queue stats
+const stats = getStats();
+```
+
+**Concurrency limits:**
+
+| Model  | Concurrent Limit |
+|--------|------------------|
+| Opus   | 3                |
+| Sonnet | 5                |
+| Haiku  | 8                |
+
+**Queue features:**
+
+- Bounded queue (max 100 tasks)
+- Per-model rate limiting
+- Task timeout: 180s individual, 10min pipeline
+- Automatic cleanup (24h TTL)
+
+## Phase Pipelining (v2.6)
+
+Prepare next phase while current phase executes:
+
+```
+Phase N execution     │ Background: Phase N+1 preparation
+──────────────────────┼──────────────────────────────────
+[Implementing...]     │ [Analyzing files for next phase]
+[Testing...]          │ [Pre-generating test cases]
+[Complete]            │ [Ready to start immediately!]
+                      ↓
+Phase N+1 starts with ZERO wait time
+```
+
+**Speed comparison:**
+
+| Mode                      | 5 Phases Total |
+|---------------------------|----------------|
+| Sequential                | ~10min         |
+| Parallel Exploration      | ~7.5min        |
+| **ULTRAWORK + Pipeline**  | **~5min**      |
+
+## PRD-to-SPEC Automation (v2.6)
+
+Auto-generate SPEC from PRD documents:
+
+```typescript
+import { parsePRD } from '@su-record/vibe/tools';
+
+// Parse PRD (supports Markdown, YAML frontmatter)
+const prd = parsePRD(prdContent, 'login-feature');
+
+// Extracted requirements with auto-generated IDs
+// REQ-login-001, REQ-login-002, ...
+```
+
+**Requirement ID System:**
+
+- Format: `REQ-{feature}-{number}`
+- Auto-deduplication
+- Priority inference from keywords (must, should, nice-to-have)
+
+## Requirements Traceability Matrix (v2.6)
+
+Track coverage from requirements to tests:
+
+```bash
+/vibe.trace "feature-name"  # Generate RTM
+```
+
+**RTM tracks:**
+
+```
+REQ-login-001 → SPEC Phase 1 → Feature Scenario 1 → login.test.ts
+REQ-login-002 → SPEC Phase 2 → Feature Scenario 3 → auth.test.ts
+```
+
+**Coverage report:**
+
+- Requirements coverage %
+- Missing implementations
+- Untested features
+- Export: Markdown, HTML
+
+## SPEC Versioning (v2.6)
+
+Automatic version management for SPEC documents:
+
+```typescript
+import { bumpVersion, generateChangelog } from '@su-record/vibe/tools';
+
+// Auto-bump version on SPEC changes
+await bumpVersion('feature-name', 'minor');
+
+// Generate changelog
+await generateChangelog('feature-name');
+```
+
+**Features:**
+
+- Semantic versioning (major.minor.patch)
+- Git tag integration
+- Automatic changelog generation
+- Baseline tagging for releases
+
+## UI Preview (`--ui`)
+
+Generate UI previews from text description or design folder:
+
+```bash
+/vibe.utils --ui "login form with email/password"
+/vibe.utils --ui ./design/dashboard/
+```
+
+**Features:**
+
+- **Gemini enabled**: Generates actual UI mockup image
+- **Gemini disabled**: ASCII art fallback
+- Supports: HTML, PNG, JPG, CSS, JSON, SVG, MD
+
+## Multi-model Orchestration
+
+**Hook-based automatic routing** - Keywords in your prompt trigger the right LLM automatically:
+
+| Keyword Pattern | Routes to | Use Case |
+|-----------------|-----------|----------|
+| `architecture`, `design` | GPT | Architecture review |
+| `UI`, `UX` | Gemini | UI/UX feedback |
+| `debugging`, `find bug` | GPT | Bug analysis |
+| `analyze code` | Gemini | Code review |
+
+**Smart Routing features:**
+- Automatic LLM selection based on task type
+- Exponential backoff retry (3 attempts, 2s → 4s → 8s)
+- Auto fallback chain: GPT ↔ Gemini (if primary fails)
+- 5-minute availability cache
+- Web search via Gemini (Google Search grounding)
+
+**API usage:**
+
+```javascript
+import('@su-record/vibe/lib/gpt').then(g => g.ask('question'))
+import('@su-record/vibe/lib/gemini').then(g => g.webSearch('search query'))
+```
+
+## Multi-LLM Research (v2.5)
+
+During `/vibe.spec`, research agents use **3 LLM perspectives** for quality assurance:
+
+```text
+┌─────────────────────────────────────────────────────────────┐
+│  MULTI-LLM PARALLEL RESEARCH                                │
+├─────────────────────────────────────────────────────────────┤
+│  best-practices-agent                                       │
+│    ├── Claude: Core patterns, anti-patterns                 │
+│    ├── GPT: Architecture patterns, code conventions         │
+│    └── Gemini: Latest trends, framework updates             │
+├─────────────────────────────────────────────────────────────┤
+│  security-advisory-agent                                    │
+│    ├── Claude: OWASP Top 10, security patterns              │
+│    ├── GPT: CVE database, vulnerability details             │
+│    └── Gemini: Latest security advisories, patches          │
+└─────────────────────────────────────────────────────────────┘
+```
+
+> vibe = Quality Assurance Framework
+
+## Race Review (v2.6.9)
+
+**Multi-LLM competitive review** - Same task runs on GPT + Gemini in parallel, results cross-validated:
+
+```bash
+/vibe.review --race              # All review types
+/vibe.review --race security     # Security only
+```
+
+**Cross-validation confidence:**
+
+| Agreement         | Priority | Action                  |
+|-------------------|----------|-------------------------|
+| Both agree (100%) | P1       | High confidence finding |
+| One model (50%)   | P2       | Needs verification      |
+
+**Output:**
+
+```text
+🏁 RACE REVIEW
+├─ GPT:    [SQL injection, XSS]
+└─ Gemini: [SQL injection, CSRF]
+         ↓
+Cross-validation:
+- SQL injection (2/2) → 🔴 P1 (100%)
+- XSS (1/2) → 🟡 P2 (50%)
+- CSRF (1/2) → 🟡 P2 (50%)
+```
+
+**ULTRAWORK default:** Race review is automatically enabled in ULTRAWORK mode.
+
+**API usage:**
+
+```javascript
+import('@su-record/vibe/tools').then(t =>
+  t.raceReview({ reviewType: 'security', code: 'CODE' })
+   .then(r => console.log(t.formatRaceResult(r)))
+)
+```
+
+## Parallel Review Agents
+
+`/vibe.review` runs 13+ agents concurrently:
+
+| Category | Agents |
+|----------|--------|
+| Security | security-reviewer, data-integrity-reviewer |
+| Performance | performance-reviewer, complexity-reviewer |
+| Architecture | architecture-reviewer, simplicity-reviewer |
+| Language | python, typescript, rails, react reviewers |
+| Context | git-history, test-coverage reviewers |
+
+**Priority:**
+- 🔴 P1 (Critical): Blocks merge
+- 🟡 P2 (Important): Recommended fix
+- 🔵 P3 (Nice-to-have): Backlog
+
+## Built-in Tools (25)
+
+| Category | Tools |
+|----------|-------|
+| **Semantic** | `find_symbol`, `find_references`, `analyze_dependency_graph` |
+| **Convention** | `analyze_complexity`, `validate_code_quality`, `check_coupling_cohesion`, `suggest_improvements`, `apply_quality_rules` |
+| **Memory** | `start_session`, `save_memory`, `auto_save_context`, `recall_memory`, `search_memories`, `search_memories_advanced`, `list_memories`, `update_memory`, `delete_memory`, `link_memories`, `prioritize_memory`, `get_memory_graph`, `create_memory_timeline`, `get_session_context`, `restore_session_context` |
+| **UI** | `preview_ui_ascii` |
+| **Time** | `get_current_time` |
+
+## Project Structure
+
+**Global install (`~/.claude/`):**
+
+```text
+~/.claude/
+├── commands/       # Slash commands (7)
+├── agents/         # Review/research agents
+├── skills/         # Auto-activated guides (7)
+└── settings.json   # Hooks settings
+```
+
+**Project-specific (`project/.claude/vibe/`):**
+
+```text
+.claude/vibe/
+├── specs/          # SPEC documents
+├── features/       # BDD scenarios
+├── rules/          # Coding rules (per stack)
+├── solutions/      # Solution archive
+├── config.json     # Project settings
+└── constitution.md # Project principles
+```
+
+## Cursor IDE Support (v2.6.8)
+
+VIBE automatically installs Cursor-compatible assets on `vibe init` and `vibe update`:
+
+**Installed locations:**
+
+| Asset | Path | Count |
+| ----- | ---- | ----- |
+| Subagents | `~/.cursor/agents/*.md` | 12 |
+| Skills | `~/.cursor/skills/*/SKILL.md` | 7 |
+| Rules Template | `~/.cursor/rules-template/*.mdc` | 5 |
+
+**12 Review Subagents with optimized model mapping:**
+
+| Agent Type | Model | Purpose |
+| ---------- | ----- | ------- |
+| security, architecture, data-integrity | `claude-4.5-sonnet-thinking` | Deep reasoning |
+| typescript, python, react, rails | `gpt-5.2-codex` | Code understanding |
+| performance, complexity, simplicity, test-coverage, git-history | `gemini-3-flash` | Fast pattern checks |
+
+**7 Skills (VIBE workflow guides):**
+
+- `vibe-spec` - SPEC creation
+- `vibe-run` - Implementation execution
+- `vibe-review` - Parallel code review
+- `vibe-analyze` - Project analysis
+- `vibe-verify` - Requirement verification
+- `vibe-reason` - Systematic reasoning
+- `vibe-ui` - UI preview utilities
+
+**5 Rules Templates:**
+
+- `typescript-standards.mdc` - Type safety, complexity limits
+- `react-patterns.mdc` - Component patterns
+- `code-quality.mdc` - General quality rules
+- `security-checklist.mdc` - OWASP checklist
+- `python-standards.mdc` - PEP8, type hints
+
+**Usage in Cursor:**
+
+```text
+# Invoke subagent
+"Use security-reviewer to check this code"
+
+# Use skill workflow
+"vibe spec login feature"
+```
+
+## Code Quality Standards
+
+| Metric | Limit |
+|--------|-------|
+| Function length | 30 lines recommended, 50 allowed |
+| Nesting depth | 3 levels max |
+| Parameters | 5 max |
+| Cyclomatic complexity | 10 max |
+
+## Usage Example
+
+```
+User: /vibe.spec "login feature"
+
+Claude: You're building a login feature! I have a few questions.
+        1. Auth method? (email/password, OAuth, Passkey)
+        2. Tech stack?
+        ...
+
+[Requirements confirmed via conversation]
+[4 parallel research agents run]
+
+✅ SPEC document created
+📄 .claude/vibe/specs/login.md
+📄 .claude/vibe/features/login.feature
+
+Next: /vibe.run "login feature" ultrawork
+```
+
+## Context7 Plugin (Recommended)
+
+Install context7 plugin for up-to-date library documentation:
+
+```bash
+/plugin install context7
+```
+
+**Benefits:**
+
+- **Skill integration**: Auto-invokes context7 for library/API questions without prompting
+- **Subagent isolation**: Docs queries run in separate context, preventing main context bloat
+- **Knowledge cutoff solution**: Access latest docs for long coding sessions
+
+**Usage:**
+
+- Automatic: Ask library questions, vibe's skill triggers context7
+- Manual: `/context7:docs <library> [query]`
+
+## Requirements
+
+- Node.js 18.0.0+
+- Claude Code
+
+## License
+
+MIT - [GitHub](https://github.com/su-record/vibe)
+
+## Vibe Setup (AI Coding)
+
+This project uses [Vibe](https://github.com/su-record/vibe) AI coding framework.
+
+### Collaborator Install
+
+```bash
+# Global install (recommended)
+npm install -g @su-record/vibe
+vibe update
+
+# Or use vibe init to setup
+vibe init
+```
+
+### Usage
+
+Use slash commands in Claude Code:
+- `/vibe.spec "feature"` - Create SPEC document
+- `/vibe.run "feature"` - Execute implementation
+
+---
+
+⭐ If this helps your workflow, consider giving it a star!