agent-capabilities
CommandMatrix of agent capabilities and their specializations.
Start a new pair programming session with AI assistance.
> /plugin marketplace add ruvnet/agentic-flow> /plugin install agentic-flow@agentic-flow
Start a new pair programming session with AI assistance.
claude-flow pair --start [options]
claude-flow pair --start
claude-flow pair --start \ --agent senior-dev \ --focus refactor \ --verify \ --threshold 0.98
claude-flow pair --start \ --mode driver \ --focus test \ --test \ --language javascript
claude-flow pair --start \ --agent debugger-expert \ --focus debug \ --review
When starting a session, the system:
1. **Selects AI Partner** - Matches expertise to your needs 2. **Configures Environment** - Sets up verification and testing 3. **Establishes Roles** - Defines driver/navigator responsibilities 4. **Loads Context** - Imports project information 5. **Begins Monitoring** - Tracks quality metrics
You write code while AI:
AI writes code while you:
Automatically alternates roles:
👥 Pair Programming Session Started ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Session ID: pair_1755021234567 Partner: senior-dev Mode: Switch (10m intervals) Focus: Implementation Language: JavaScript Verification: ✅ Enabled (0.95 threshold) Testing: ✅ Auto-run on save Review: ✅ Continuous Current Role: DRIVER (you) Navigator: senior-dev is ready... 📝 Workspace: /workspaces/my-project 📊 Truth Score: 0.972 ✅ 🧪 Test Coverage: 84% Type /help for commands or start coding...
Start sessions in background for long-running collaboration:
# Start in background claude-flow pair --start --background # Monitor session claude-flow pair status # View session output claude-flow pair output session_id # End background session claude-flow pair --end session_id
claude-flow pair --start --git --auto-commit
claude-flow pair --start --ci --non-interactive
claude-flow pair --start --ide vscode
1. **Clear Goals** - Define session objectives 2. **Appropriate Mode** - Choose based on task 3. **Enable Verification** - For critical code 4. **Regular Testing** - Maintain quality 5. **Session Notes** - Document decisions
Production-ready AI agent orchestration with 66 self-learning agents, 213 MCP tools, and autonomous multi-agent swarms.
Matrix of agent capabilities and their specializations.
Complete guide to all 54 available agent types in Claude Flow.
Reviewed all command files in `.claude/commands/analysis/` directory to ensure proper usage of: - `mcp__claude-flow__*` tools…
Analyze performance bottlenecks in swarm operations and suggest optimizations.
Implement tasks incrementally — build, test, verify, commit. Add "auto" to run the whole plan in one approved pass.