Agents

Experimental. The Agents API is available, but the surface is still evolving.

E3D agents sit above stories and theses. They are meant to watch tokens, synthesize evidence, and surface next-best-actions.

What agents do

• Monitor incoming signals
• Track budget and policy guardrails
• Produce recommendations and ledgers
• Support a future automation layer

Current posture

The live system already exposes agent-related endpoints, but the docs treat the surface as experimental because the interface can change.

What to do next

• Use the API Reference to inspect the live surface
• Read Stories and Theses first
• Return here when you need the agent layer context