Agents

An agent is the smallest independently executable unit in UnderOcean. It bundles together everything needed to have a model respond to input:

• A system prompt (versioned — editing it creates a new agent version)
• A provider configuration (which model, via which provider)
• Generation parameters: temperature, max tokens
• Optional reasoning effort (none | low | medium | high) — when set on a model that supports extended thinking/reasoning, its reasoning trace is streamed and persisted alongside the answer; see Executions & Streaming
• Optional tool assignments (see MCP)
• Optional memory configuration (see Context Management)
• Optional knowledge base access for Agentic-RAG

Versioning

Agents are versioned rather than mutated in place. Updating the system prompt creates a new version; past versions remain available for rollback and for correlating historical executions with the prompt that produced them.

Execution

An agent is executed by creating an execution resource. Every execution runs on the same orchestration engine, so streaming, checkpointing, and reasoning-step visibility apply uniformly whether you're running a single agent or a multi-step flow. See Executions & Streaming.

Guardrails

Agents can be configured with input/output safety controls (e.g. blocked terms, prompt-injection mitigations). User input to agents always passes through prompt-injection protection before reaching the model — this is a security boundary, not an optional feature.

Related

• Flows — orchestrate multiple agents together
• Providers — how model access is configured
• Guides → Build your first agent