diff --git a/.agentic-qe/config.yaml b/.agentic-qe/config.yaml
new file mode 100644
index 0000000..b2cd779
--- /dev/null
+++ b/.agentic-qe/config.yaml
@@ -0,0 +1,79 @@
+# Agentic QE v3 Configuration
+# Generated by aqe init
+# 2026-02-19T17:38:27.216Z
+#
+# NOTE: Your customizations are PRESERVED when you run "aqe init" again.
+# You do NOT need to re-run "aqe init" after editing this file - changes
+# take effect immediately. The following settings are merged on reinstall:
+#   - domains.enabled (custom domains like visual-accessibility)
+#   - domains.disabled
+#   - learning.enabled
+#   - hooks.* preferences
+#   - workers.enabled
+#   - agents.maxConcurrent and defaultTimeout
+
+version: "3.6.13"
+
+project:
+  name: "cf-devpod"
+  root: "/home/user/cf-devpod"
+  type: "single"
+
+learning:
+  enabled: true
+  embeddingModel: "transformer"
+  hnswConfig:
+    M: 8
+    efConstruction: 100
+    efSearch: 50
+  qualityThreshold: 0.5
+  promotionThreshold: 2
+  pretrainedPatterns: true
+
+routing:
+  mode: "ml"
+  confidenceThreshold: 0.7
+  feedbackEnabled: true
+
+workers:
+  enabled:
+    - "pattern-consolidator"
+  intervals:
+    pattern-consolidator: 1800000
+    coverage-gap-scanner: 3600000
+    flaky-test-detector: 7200000
+    routing-accuracy-monitor: 900000
+  maxConcurrent: 2
+  daemonAutoStart: true
+
+hooks:
+  claudeCode: true
+  preCommit: false
+  ciIntegration: false
+
+skills:
+  install: true
+  installV2: true
+  installV3: true
+  overwrite: false
+
+domains:
+  enabled:
+    - "test-generation"
+    - "test-execution"
+    - "coverage-analysis"
+    - "quality-assessment"
+    - "defect-intelligence"
+    - "requirements-validation"
+    - "code-intelligence"
+    - "security-compliance"
+    - "contract-testing"
+    - "visual-accessibility"
+    - "chaos-resilience"
+    - "learning-optimization"
+    - "enterprise-integration"
+  disabled:
+
+agents:
+  maxConcurrent: 5
+  defaultTimeout: 60000
diff --git a/.agentic-qe/data/learning-config.json b/.agentic-qe/data/learning-config.json
new file mode 100644
index 0000000..d371d2e
--- /dev/null
+++ b/.agentic-qe/data/learning-config.json
@@ -0,0 +1,13 @@
+{
+  "embeddingModel": "transformer",
+  "hnswConfig": {
+    "M": 8,
+    "efConstruction": 100,
+    "efSearch": 50
+  },
+  "qualityThreshold": 0.5,
+  "promotionThreshold": 2,
+  "databasePath": "/home/user/cf-devpod/.agentic-qe/data/memory.db",
+  "hnswIndexPath": "/home/user/cf-devpod/.agentic-qe/data/hnsw/index.bin",
+  "initialized": "2026-02-19T17:38:24.021Z"
+}
\ No newline at end of file
diff --git a/.agentic-qe/data/patterns/index.json b/.agentic-qe/data/patterns/index.json
new file mode 100644
index 0000000..091a3c0
--- /dev/null
+++ b/.agentic-qe/data/patterns/index.json
@@ -0,0 +1,5 @@
+{
+  "version": "3.6.13",
+  "domains": [],
+  "loadedAt": "2026-02-19T17:38:24.022Z"
+}
\ No newline at end of file
diff --git a/.agentic-qe/memory.db-shm b/.agentic-qe/memory.db-shm
new file mode 100644
index 0000000..c16dffc
Binary files /dev/null and b/.agentic-qe/memory.db-shm differ
diff --git a/.agentic-qe/workers/daemon.pid b/.agentic-qe/workers/daemon.pid
new file mode 100644
index 0000000..673bb65
--- /dev/null
+++ b/.agentic-qe/workers/daemon.pid
@@ -0,0 +1 @@
+2575
\ No newline at end of file
diff --git a/.agentic-qe/workers/pattern-consolidator.json b/.agentic-qe/workers/pattern-consolidator.json
new file mode 100644
index 0000000..b96a347
--- /dev/null
+++ b/.agentic-qe/workers/pattern-consolidator.json
@@ -0,0 +1,8 @@
+{
+  "name": "pattern-consolidator",
+  "enabled": true,
+  "interval": 1800000,
+  "projectRoot": "/home/user/cf-devpod",
+  "dataDir": "/home/user/cf-devpod/.agentic-qe/data",
+  "createdAt": "2026-02-19T17:38:25.645Z"
+}
\ No newline at end of file
diff --git a/.agentic-qe/workers/registry.json b/.agentic-qe/workers/registry.json
new file mode 100644
index 0000000..33bf1e4
--- /dev/null
+++ b/.agentic-qe/workers/registry.json
@@ -0,0 +1,15 @@
+{
+  "version": "3.6.13",
+  "maxConcurrent": 2,
+  "workers": {
+    "pattern-consolidator": {
+      "name": "pattern-consolidator",
+      "enabled": true,
+      "interval": 1800000,
+      "lastRun": null,
+      "status": "pending"
+    }
+  },
+  "createdAt": "2026-02-19T17:38:25.645Z",
+  "daemonPid": null
+}
\ No newline at end of file
diff --git a/.agentic-qe/workers/start-daemon.cjs b/.agentic-qe/workers/start-daemon.cjs
new file mode 100644
index 0000000..198907b
--- /dev/null
+++ b/.agentic-qe/workers/start-daemon.cjs
@@ -0,0 +1,56 @@
+#!/usr/bin/env node
+// AQE v3 Worker Daemon Startup Script (cross-platform)
+// Generated by aqe init
+// Starts the MCP server with background workers enabled
+
+const { existsSync, readFileSync, writeFileSync, appendFileSync } = require('fs');
+const { join } = require('path');
+const { spawn } = require('child_process');
+
+const projectRoot = join(__dirname, '..', '..');
+const workersDir = join(projectRoot, '.agentic-qe', 'workers');
+const pidFile = join(workersDir, 'daemon.pid');
+const logFile = join(workersDir, 'daemon.log');
+
+// Check if already running
+if (existsSync(pidFile)) {
+  const pid = parseInt(readFileSync(pidFile, 'utf-8').trim(), 10);
+  try { process.kill(pid, 0); console.log('Daemon already running (PID: ' + pid + ')'); process.exit(0); } catch {}
+}
+
+const ts = new Date().toISOString();
+appendFileSync(logFile, '[' + ts + '] Starting AQE v3 Worker Daemon...\n');
+
+// Find the best way to run aqe-mcp
+const candidates = [
+  join(projectRoot, 'node_modules', '.bin', 'aqe-mcp'),
+  join(projectRoot, 'node_modules', 'agentic-qe', 'v3', 'dist', 'mcp', 'bundle.js'),
+];
+
+let mcpCmd, mcpArgs;
+const binCandidate = candidates.find(c => existsSync(c));
+
+if (binCandidate && binCandidate.endsWith('bundle.js')) {
+  mcpCmd = process.execPath;
+  mcpArgs = [binCandidate];
+} else if (binCandidate) {
+  mcpCmd = binCandidate;
+  mcpArgs = [];
+} else {
+  mcpCmd = process.platform === 'win32' ? 'npx.cmd' : 'npx';
+  mcpArgs = ['--yes', 'agentic-qe', 'mcp'];
+}
+
+appendFileSync(logFile, '[' + ts + '] Using: ' + mcpCmd + ' ' + mcpArgs.join(' ') + '\n');
+
+const env = { ...process.env, AQE_PROJECT_ROOT: projectRoot, AQE_LEARNING_ENABLED: 'true', AQE_WORKERS_ENABLED: 'true', AQE_HTTP_PORT: '0' };
+const child = spawn(mcpCmd, mcpArgs, { cwd: projectRoot, env, detached: true, stdio: ['ignore', 'pipe', 'pipe'] });
+child.stdout.on('data', d => appendFileSync(logFile, d));
+child.stderr.on('data', d => appendFileSync(logFile, d));
+child.unref();
+
+writeFileSync(pidFile, String(child.pid));
+appendFileSync(logFile, '[' + ts + '] Daemon started with PID: ' + child.pid + '\n');
+console.log('AQE v3 Worker Daemon started (PID: ' + child.pid + ')');
+console.log('Log file: ' + logFile);
+console.log('To stop: node ' + join(workersDir, 'stop-daemon.cjs'));
diff --git a/.agentic-qe/workers/stop-daemon.cjs b/.agentic-qe/workers/stop-daemon.cjs
new file mode 100644
index 0000000..fcc303e
--- /dev/null
+++ b/.agentic-qe/workers/stop-daemon.cjs
@@ -0,0 +1,23 @@
+#!/usr/bin/env node
+// AQE v3 Worker Daemon Stop Script (cross-platform)
+// Generated by aqe init
+
+const { existsSync, readFileSync, unlinkSync } = require('fs');
+const { join } = require('path');
+
+const workersDir = join(__dirname);
+const pidFile = join(workersDir, 'daemon.pid');
+
+if (!existsSync(pidFile)) { console.log('No PID file found - daemon may not be running'); process.exit(0); }
+
+const pid = parseInt(readFileSync(pidFile, 'utf-8').trim(), 10);
+try {
+  process.kill(pid, 0); // Check if alive
+  console.log('Stopping daemon (PID: ' + pid + ')...');
+  process.kill(pid, 'SIGTERM');
+  try { unlinkSync(pidFile); } catch {}
+  console.log('Daemon stopped');
+} catch {
+  console.log('Daemon not running (stale PID file)');
+  try { unlinkSync(pidFile); } catch {}
+}
diff --git a/CLAUDE.md b/CLAUDE.md
new file mode 100644
index 0000000..804b92c
--- /dev/null
+++ b/CLAUDE.md
@@ -0,0 +1,352 @@
+# Claude Code Configuration - Claude Flow V3
+
+## Behavioral Rules (Always Enforced)
+
+- Do what has been asked; nothing more, nothing less
+- NEVER create files unless they're absolutely necessary for achieving your goal
+- ALWAYS prefer editing an existing file to creating a new one
+- NEVER proactively create documentation files (*.md) or README files unless explicitly requested
+- NEVER save working files, text/mds, or tests to the root folder
+- Never continuously check status after spawning a swarm — wait for results
+- ALWAYS read a file before editing it
+- NEVER commit secrets, credentials, or .env files
+
+## File Organization
+
+- NEVER save to root folder — use the directories below
+- Use `/src` for source code files
+- Use `/tests` for test files
+- Use `/docs` for documentation and markdown files
+- Use `/config` for configuration files
+- Use `/scripts` for utility scripts
+- Use `/examples` for example code
+
+## Project Architecture
+
+- Follow Domain-Driven Design with bounded contexts
+- Keep files under 500 lines
+- Use typed interfaces for all public APIs
+- Prefer TDD London School (mock-first) for new code
+- Use event sourcing for state changes
+- Ensure input validation at system boundaries
+
+### Project Config
+
+- **Topology**: hierarchical-mesh
+- **Max Agents**: 15
+- **Memory**: hybrid
+- **HNSW**: Enabled
+- **Neural**: Enabled
+
+## Build & Test
+
+```bash
+# Build
+npm run build
+
+# Test
+npm test
+
+# Lint
+npm run lint
+```
+
+- ALWAYS run tests after making code changes
+- ALWAYS verify build succeeds before committing
+
+## Security Rules
+
+- NEVER hardcode API keys, secrets, or credentials in source files
+- NEVER commit .env files or any file containing secrets
+- Always validate user input at system boundaries
+- Always sanitize file paths to prevent directory traversal
+- Run `npx @claude-flow/cli@latest security scan` after security-related changes
+
+## Concurrency: 1 MESSAGE = ALL RELATED OPERATIONS
+
+- All operations MUST be concurrent/parallel in a single message
+- Use Claude Code's Task tool for spawning agents, not just MCP
+- ALWAYS batch ALL todos in ONE TodoWrite call (5-10+ minimum)
+- ALWAYS spawn ALL agents in ONE message with full instructions via Task tool
+- ALWAYS batch ALL file reads/writes/edits in ONE message
+- ALWAYS batch ALL Bash commands in ONE message
+
+## Swarm Orchestration
+
+- MUST initialize the swarm using CLI tools when starting complex tasks
+- MUST spawn concurrent agents using Claude Code's Task tool
+- Never use CLI tools alone for execution — Task tool agents do the actual work
+- MUST call CLI tools AND Task tool in ONE message for complex work
+
+### 3-Tier Model Routing (ADR-026)
+
+| Tier | Handler | Latency | Cost | Use Cases |
+|------|---------|---------|------|-----------|
+| **1** | Agent Booster (WASM) | <1ms | $0 | Simple transforms (var→const, add types) — Skip LLM |
+| **2** | Haiku | ~500ms | $0.0002 | Simple tasks, low complexity (<30%) |
+| **3** | Sonnet/Opus | 2-5s | $0.003-0.015 | Complex reasoning, architecture, security (>30%) |
+
+- Always check for `[AGENT_BOOSTER_AVAILABLE]` or `[TASK_MODEL_RECOMMENDATION]` before spawning agents
+- Use Edit tool directly when `[AGENT_BOOSTER_AVAILABLE]`
+
+## Swarm Configuration & Anti-Drift
+
+- ALWAYS use hierarchical topology for coding swarms
+- Keep maxAgents at 6-8 for tight coordination
+- Use specialized strategy for clear role boundaries
+- Use `raft` consensus for hive-mind (leader maintains authoritative state)
+- Run frequent checkpoints via `post-task` hooks
+- Keep shared memory namespace for all agents
+
+```bash
+npx @claude-flow/cli@latest swarm init --topology hierarchical --max-agents 8 --strategy specialized
+```
+
+## Swarm Execution Rules
+
+- ALWAYS use `run_in_background: true` for all agent Task calls
+- ALWAYS put ALL agent Task calls in ONE message for parallel execution
+- After spawning, STOP — do NOT add more tool calls or check status
+- Never poll TaskOutput or check swarm status — trust agents to return
+- When agent results arrive, review ALL results before proceeding
+
+## V3 CLI Commands
+
+### Core Commands
+
+| Command | Subcommands | Description |
+|---------|-------------|-------------|
+| `init` | 4 | Project initialization |
+| `agent` | 8 | Agent lifecycle management |
+| `swarm` | 6 | Multi-agent swarm coordination |
+| `memory` | 11 | AgentDB memory with HNSW search |
+| `task` | 6 | Task creation and lifecycle |
+| `session` | 7 | Session state management |
+| `hooks` | 17 | Self-learning hooks + 12 workers |
+| `hive-mind` | 6 | Byzantine fault-tolerant consensus |
+
+### Quick CLI Examples
+
+```bash
+npx @claude-flow/cli@latest init --wizard
+npx @claude-flow/cli@latest agent spawn -t coder --name my-coder
+npx @claude-flow/cli@latest swarm init --v3-mode
+npx @claude-flow/cli@latest memory search --query "authentication patterns"
+npx @claude-flow/cli@latest doctor --fix
+```
+
+## Available Agents (60+ Types)
+
+### Core Development
+`coder`, `reviewer`, `tester`, `planner`, `researcher`
+
+### Specialized
+`security-architect`, `security-auditor`, `memory-specialist`, `performance-engineer`
+
+### Swarm Coordination
+`hierarchical-coordinator`, `mesh-coordinator`, `adaptive-coordinator`
+
+### GitHub & Repository
+`pr-manager`, `code-review-swarm`, `issue-tracker`, `release-manager`
+
+### SPARC Methodology
+`sparc-coord`, `sparc-coder`, `specification`, `pseudocode`, `architecture`
+
+## Memory Commands Reference
+
+```bash
+# Store (REQUIRED: --key, --value; OPTIONAL: --namespace, --ttl, --tags)
+npx @claude-flow/cli@latest memory store --key "pattern-auth" --value "JWT with refresh" --namespace patterns
+
+# Search (REQUIRED: --query; OPTIONAL: --namespace, --limit, --threshold)
+npx @claude-flow/cli@latest memory search --query "authentication patterns"
+
+# List (OPTIONAL: --namespace, --limit)
+npx @claude-flow/cli@latest memory list --namespace patterns --limit 10
+
+# Retrieve (REQUIRED: --key; OPTIONAL: --namespace)
+npx @claude-flow/cli@latest memory retrieve --key "pattern-auth" --namespace patterns
+```
+
+## Quick Setup
+
+```bash
+claude mcp add claude-flow -- npx -y @claude-flow/cli@latest
+npx @claude-flow/cli@latest daemon start
+npx @claude-flow/cli@latest doctor --fix
+```
+
+## Claude Code vs CLI Tools
+
+- Claude Code's Task tool handles ALL execution: agents, file ops, code generation, git
+- CLI tools handle coordination via Bash: swarm init, memory, hooks, routing
+- NEVER use CLI tools as a substitute for Task tool agents
+
+## Support
+
+- Documentation: https://github.com/ruvnet/claude-flow
+- Issues: https://github.com/ruvnet/claude-flow/issues
+
+
+## Agentic QE v3
+
+This project uses **Agentic QE v3** - a Domain-Driven Quality Engineering platform with 13 bounded contexts, ReasoningBank learning, HNSW vector search, and Agent Teams coordination (ADR-064).
+
+---
+
+### CRITICAL POLICIES
+
+#### Integrity Rule (ABSOLUTE)
+- NO shortcuts, fake data, or false claims
+- ALWAYS implement properly, verify before claiming success
+- ALWAYS use real database queries for integration tests
+- ALWAYS run actual tests, not assume they pass
+
+**We value the quality we deliver to our users.**
+
+#### Test Execution
+- NEVER run `npm test` without `--run` flag (watch mode risk)
+- Use: `npm test -- --run`, `npm run test:unit`, `npm run test:integration` when available
+
+#### Data Protection
+- NEVER run `rm -f` on `.agentic-qe/` or `*.db` files without confirmation
+- ALWAYS backup before database operations
+
+#### Git Operations
+- NEVER auto-commit/push without explicit user request
+- ALWAYS wait for user confirmation before git operations
+
+---
+
+### Quick Reference
+
+```bash
+# Run tests
+npm test -- --run
+
+# Check quality
+aqe quality assess
+
+# Generate tests
+aqe test generate <file>
+
+# Coverage analysis
+aqe coverage <path>
+```
+
+### Using AQE MCP Tools
+
+AQE exposes tools via MCP with the `mcp__agentic-qe__` prefix. You MUST call `fleet_init` before any other tool.
+
+#### 1. Initialize the Fleet (required first step)
+
+```typescript
+mcp__agentic-qe__fleet_init({
+  topology: "hierarchical",
+  maxAgents: 15,
+  memoryBackend: "hybrid"
+})
+```
+
+#### 2. Generate Tests
+
+```typescript
+mcp__agentic-qe__test_generate_enhanced({
+  targetPath: "src/services/auth.ts",
+  framework: "vitest",
+  strategy: "boundary-value"
+})
+```
+
+#### 3. Analyze Coverage
+
+```typescript
+mcp__agentic-qe__coverage_analyze_sublinear({
+  paths: ["src/"],
+  threshold: 80
+})
+```
+
+#### 4. Assess Quality
+
+```typescript
+mcp__agentic-qe__quality_assess({
+  scope: "full",
+  includeMetrics: true
+})
+```
+
+#### 5. Store and Query Patterns (with learning persistence)
+
+```typescript
+// Store a learned pattern
+mcp__agentic-qe__memory_store({
+  key: "patterns/coverage-gap/{timestamp}",
+  namespace: "learning",
+  value: {
+    pattern: "...",
+    confidence: 0.95,
+    type: "coverage-gap",
+    metadata: { /* domain-specific */ }
+  },
+  persist: true
+})
+
+// Query stored patterns
+mcp__agentic-qe__memory_query({
+  pattern: "patterns/*",
+  namespace: "learning",
+  limit: 10
+})
+```
+
+#### 6. Orchestrate Multi-Agent Tasks
+
+```typescript
+mcp__agentic-qe__task_orchestrate({
+  task: "Full quality assessment of auth module",
+  domains: ["test-generation", "coverage-analysis", "security-compliance"],
+  parallel: true
+})
+```
+
+### MCP Tool Reference
+
+| Tool | Description |
+|------|-------------|
+| `fleet_init` | Initialize QE fleet (MUST call first) |
+| `fleet_status` | Get fleet health and agent status |
+| `agent_spawn` | Spawn specialized QE agent |
+| `test_generate_enhanced` | AI-powered test generation |
+| `test_execute_parallel` | Parallel test execution with retry |
+| `task_orchestrate` | Orchestrate multi-agent QE tasks |
+| `coverage_analyze_sublinear` | O(log n) coverage analysis |
+| `quality_assess` | Quality gate evaluation |
+| `memory_store` | Store patterns with namespace + persist |
+| `memory_query` | Query patterns by namespace/pattern |
+| `security_scan_comprehensive` | SAST/DAST scanning |
+
+### Configuration
+
+- **Enabled Domains**: test-generation, test-execution, coverage-analysis, quality-assessment, defect-intelligence, requirements-validation (+7 more)
+- **Learning**: Enabled (transformer embeddings)
+- **Max Concurrent Agents**: 5
+- **Background Workers**: pattern-consolidator
+
+### V3 QE Agents
+
+V3 QE agents are in `.claude/agents/v3/`. Use with Task tool:
+
+```javascript
+Task({ prompt: "Generate tests", subagent_type: "qe-test-architect", run_in_background: true })
+Task({ prompt: "Find coverage gaps", subagent_type: "qe-coverage-specialist", run_in_background: true })
+Task({ prompt: "Security audit", subagent_type: "qe-security-scanner", run_in_background: true })
+```
+
+### Data Storage
+
+- **Memory Backend**: `.agentic-qe/memory.db` (SQLite)
+- **Configuration**: `.agentic-qe/config.yaml`
+
+---
+*Generated by AQE v3 init - 2026-02-19T17:38:27.211Z*
diff --git a/CLAUDE.md.backup b/CLAUDE.md.backup
new file mode 100644
index 0000000..27c6e8f
--- /dev/null
+++ b/CLAUDE.md.backup
@@ -0,0 +1,188 @@
+# Claude Code Configuration - Claude Flow V3
+
+## Behavioral Rules (Always Enforced)
+
+- Do what has been asked; nothing more, nothing less
+- NEVER create files unless they're absolutely necessary for achieving your goal
+- ALWAYS prefer editing an existing file to creating a new one
+- NEVER proactively create documentation files (*.md) or README files unless explicitly requested
+- NEVER save working files, text/mds, or tests to the root folder
+- Never continuously check status after spawning a swarm — wait for results
+- ALWAYS read a file before editing it
+- NEVER commit secrets, credentials, or .env files
+
+## File Organization
+
+- NEVER save to root folder — use the directories below
+- Use `/src` for source code files
+- Use `/tests` for test files
+- Use `/docs` for documentation and markdown files
+- Use `/config` for configuration files
+- Use `/scripts` for utility scripts
+- Use `/examples` for example code
+
+## Project Architecture
+
+- Follow Domain-Driven Design with bounded contexts
+- Keep files under 500 lines
+- Use typed interfaces for all public APIs
+- Prefer TDD London School (mock-first) for new code
+- Use event sourcing for state changes
+- Ensure input validation at system boundaries
+
+### Project Config
+
+- **Topology**: hierarchical-mesh
+- **Max Agents**: 15
+- **Memory**: hybrid
+- **HNSW**: Enabled
+- **Neural**: Enabled
+
+## Build & Test
+
+```bash
+# Build
+npm run build
+
+# Test
+npm test
+
+# Lint
+npm run lint
+```
+
+- ALWAYS run tests after making code changes
+- ALWAYS verify build succeeds before committing
+
+## Security Rules
+
+- NEVER hardcode API keys, secrets, or credentials in source files
+- NEVER commit .env files or any file containing secrets
+- Always validate user input at system boundaries
+- Always sanitize file paths to prevent directory traversal
+- Run `npx @claude-flow/cli@latest security scan` after security-related changes
+
+## Concurrency: 1 MESSAGE = ALL RELATED OPERATIONS
+
+- All operations MUST be concurrent/parallel in a single message
+- Use Claude Code's Task tool for spawning agents, not just MCP
+- ALWAYS batch ALL todos in ONE TodoWrite call (5-10+ minimum)
+- ALWAYS spawn ALL agents in ONE message with full instructions via Task tool
+- ALWAYS batch ALL file reads/writes/edits in ONE message
+- ALWAYS batch ALL Bash commands in ONE message
+
+## Swarm Orchestration
+
+- MUST initialize the swarm using CLI tools when starting complex tasks
+- MUST spawn concurrent agents using Claude Code's Task tool
+- Never use CLI tools alone for execution — Task tool agents do the actual work
+- MUST call CLI tools AND Task tool in ONE message for complex work
+
+### 3-Tier Model Routing (ADR-026)
+
+| Tier | Handler | Latency | Cost | Use Cases |
+|------|---------|---------|------|-----------|
+| **1** | Agent Booster (WASM) | <1ms | $0 | Simple transforms (var→const, add types) — Skip LLM |
+| **2** | Haiku | ~500ms | $0.0002 | Simple tasks, low complexity (<30%) |
+| **3** | Sonnet/Opus | 2-5s | $0.003-0.015 | Complex reasoning, architecture, security (>30%) |
+
+- Always check for `[AGENT_BOOSTER_AVAILABLE]` or `[TASK_MODEL_RECOMMENDATION]` before spawning agents
+- Use Edit tool directly when `[AGENT_BOOSTER_AVAILABLE]`
+
+## Swarm Configuration & Anti-Drift
+
+- ALWAYS use hierarchical topology for coding swarms
+- Keep maxAgents at 6-8 for tight coordination
+- Use specialized strategy for clear role boundaries
+- Use `raft` consensus for hive-mind (leader maintains authoritative state)
+- Run frequent checkpoints via `post-task` hooks
+- Keep shared memory namespace for all agents
+
+```bash
+npx @claude-flow/cli@latest swarm init --topology hierarchical --max-agents 8 --strategy specialized
+```
+
+## Swarm Execution Rules
+
+- ALWAYS use `run_in_background: true` for all agent Task calls
+- ALWAYS put ALL agent Task calls in ONE message for parallel execution
+- After spawning, STOP — do NOT add more tool calls or check status
+- Never poll TaskOutput or check swarm status — trust agents to return
+- When agent results arrive, review ALL results before proceeding
+
+## V3 CLI Commands
+
+### Core Commands
+
+| Command | Subcommands | Description |
+|---------|-------------|-------------|
+| `init` | 4 | Project initialization |
+| `agent` | 8 | Agent lifecycle management |
+| `swarm` | 6 | Multi-agent swarm coordination |
+| `memory` | 11 | AgentDB memory with HNSW search |
+| `task` | 6 | Task creation and lifecycle |
+| `session` | 7 | Session state management |
+| `hooks` | 17 | Self-learning hooks + 12 workers |
+| `hive-mind` | 6 | Byzantine fault-tolerant consensus |
+
+### Quick CLI Examples
+
+```bash
+npx @claude-flow/cli@latest init --wizard
+npx @claude-flow/cli@latest agent spawn -t coder --name my-coder
+npx @claude-flow/cli@latest swarm init --v3-mode
+npx @claude-flow/cli@latest memory search --query "authentication patterns"
+npx @claude-flow/cli@latest doctor --fix
+```
+
+## Available Agents (60+ Types)
+
+### Core Development
+`coder`, `reviewer`, `tester`, `planner`, `researcher`
+
+### Specialized
+`security-architect`, `security-auditor`, `memory-specialist`, `performance-engineer`
+
+### Swarm Coordination
+`hierarchical-coordinator`, `mesh-coordinator`, `adaptive-coordinator`
+
+### GitHub & Repository
+`pr-manager`, `code-review-swarm`, `issue-tracker`, `release-manager`
+
+### SPARC Methodology
+`sparc-coord`, `sparc-coder`, `specification`, `pseudocode`, `architecture`
+
+## Memory Commands Reference
+
+```bash
+# Store (REQUIRED: --key, --value; OPTIONAL: --namespace, --ttl, --tags)
+npx @claude-flow/cli@latest memory store --key "pattern-auth" --value "JWT with refresh" --namespace patterns
+
+# Search (REQUIRED: --query; OPTIONAL: --namespace, --limit, --threshold)
+npx @claude-flow/cli@latest memory search --query "authentication patterns"
+
+# List (OPTIONAL: --namespace, --limit)
+npx @claude-flow/cli@latest memory list --namespace patterns --limit 10
+
+# Retrieve (REQUIRED: --key; OPTIONAL: --namespace)
+npx @claude-flow/cli@latest memory retrieve --key "pattern-auth" --namespace patterns
+```
+
+## Quick Setup
+
+```bash
+claude mcp add claude-flow -- npx -y @claude-flow/cli@latest
+npx @claude-flow/cli@latest daemon start
+npx @claude-flow/cli@latest doctor --fix
+```
+
+## Claude Code vs CLI Tools
+
+- Claude Code's Task tool handles ALL execution: agents, file ops, code generation, git
+- CLI tools handle coordination via Bash: swarm init, memory, hooks, routing
+- NEVER use CLI tools as a substitute for Task tool agents
+
+## Support
+
+- Documentation: https://github.com/ruvnet/claude-flow
+- Issues: https://github.com/ruvnet/claude-flow/issues