| Agent | The React tree in src/agent/agent.tsx, and everything it renders |
| Agent level | What the agent renders before the model has read a skill. It is in every request |
| App message | A message between the agent and the page or the host, with no model involved (App messages) |
| Artifact | What curb build produces: a project that starts with no install (Self-hosting) |
| Client | The caller’s page. What it says arrives as client.*, and it is never proof of identity |
| Context | What the model reads in a request (What the model receives) |
| Diagnostic | A problem curb check reports, with a stable code and a fix (Diagnostics) |
| Experiment | Variants of an agent, assigned per call (Goals and experiments) |
| Fake | The result a tool gives in a test instead of running (Testing) |
| Goal | What counts as a successful call (Goals and experiments) |
| Host | What runs an agent and gives it calls (Host protocol) |
| Procedure | An ordered flow of steps (Step-by-step procedures) |
| Provider | The LLM, speech-to-text or text-to-speech an agent runs on (Providers) |
| Runner | One process that serves calls for a host |
| Scenario | A whole conversation kept as a test (Testing) |
| Skill | A whole area of work the model opens when it needs it (Agents and skills) |
| Store | State that lasts the whole call (State and conditional tools) |
| Tool | A single action the model can call (Tools) |
| Turn | One thing the caller says and everything the agent does in answer |