Start with these guides
Connect your first Claw, then expand by topic
If this is your first time, begin with quickstart. Once your first conversation works, expand via Agent, Skill, MCP, Trigger and Memory guides.
All topics
Pick a topic to start reading
Getting Started
Configure Agents
- Agent Configuration Create, edit, and manage Agents; understand system prompts, tool whitelists, built-in tools, prompt boosts, and visibility.
- LLM & Credentials Configure Anthropic / OpenAI / OpenRouter / custom providers; assign dedicated credentials for Codex / Claude Code tools; set up multimodal reading and local reclaude.
Extensions
- Skill Overview Skills live on the Claw host filesystem. This page only tells you where they live and who manages them β ask the General Assistant to install new ones.
- MCP Configuration Configure MCP Servers on Claw and mount tool sets per session; understand the defaultConfig and manageDoc fields.
- Trigger Use cron expressions or webhooks to have Agents start automatically; integrate with GitHub Actions and other systems via TypeScript / Python / C# SDKs.
Multi-machine & Subagents
- Claw Deploy Install Claw on Windows / macOS / Linux hosts. Prefer the "Add Claw" button inside Client; the scripted path is for batch and automation use cases.
- Subagent Orchestration Use Claude Code / Codex / Orchestrator subagents to break down long tasks; understand foreground/background modes, resumability, message backflow, and credential inheritance.
Advanced
- Cross-conversation Memory Use the clue tree / dream governance / four memory types to give Agents context continuity across multiple conversations, projects, and time.
- Desktop Power Bash attached desktop sessions, modified files panel, Command Monitor, conversation forking, slash commands, and cross-Claw active session toggle.
- Mobile & Push Onboarding, task board, Agent message center, push notifications, and offline background interactions.