Visual Maps
These 8 diagrams together explain how OpenClaw works. Read them in order if you’re new — each builds on the previous.
| # | Map | What it answers |
|---|---|---|
| 1 | Master Architecture | ”Where does the data live and what talks to what?“ |
| 2 | Chat → Execution Sequence | ”What happens between my prompt and the response?“ |
| 3 | Hook Lifecycle | ”What runs around every tool call?“ |
| 4 | MCP Server Catalog | ”What are MCPs and which ones are wired in?“ |
| 5 | Model Routing | ”When does Opus vs Sonnet vs Haiku fire?“ |
| 6 | Memory Pipeline | ”How does memory + vector search work?“ |
| 7 | Agent Dispatch Sequence | ”How does Opus delegate execution?“ |
| 8 | Skill Trigger Keyword Map | ”How do slash commands and auto-triggers work?” |
For deeper-tier system maps (62 internal Tier 2/3 maps tracked separately), browse the vault at /vault/wiki/system-map/.