Use these guides after configuring the core agent API:
- Tools & Security — validation, approvals, retries, errors, secrets, and sandbox boundaries.
- Sessions & Memory — fetch routes, channels, streaming, session ordering, and persistence.
- Testing — deterministic model, channel, HTTP, MCP, lifecycle, and tool tests.
- Terminal UI — add the optional
@xsaf/cliinteractive terminal interface.
For method signatures and configuration rules, start with the agent API reference.