Learn
AgentField Documentation
The AI backend. Build, deploy, and govern AI agents like APIs.
The AI backend. Build, deploy, and govern AI agents like APIs.
curl -sSf https://agentfield.ai/get | sh/
Fastest way in: one prompt, full backend
After install, type /agentfield in Claude Code — or describe your system in plain English. Stack is live under docker compose in one shot.
Try it →
from agentfield import Agent, AIConfig
from pydantic import BaseModel
app = Agent("demo", ai_config=AIConfig(model="anthropic/claude-sonnet-4-20250514"))
class Decision(BaseModel):
action: str
@app.reasoner()
async def route(text: str) -> dict:
out = await app.ai(
system="Pick one action: summarize | escalate | done.",
user=text,
schema=Decision,
)
return out.model_dump()
app.run()import { Agent } from "@agentfield/sdk";
import { z } from "zod";
const agent = new Agent({
nodeId: "demo",
aiConfig: { provider: "anthropic", model: "claude-sonnet-4-20250514" },
});
agent.reasoner("route", async (ctx) => {
return await ctx.ai(ctx.input.text, {
system: "Pick one action: summarize | escalate | done.",
schema: z.object({ action: z.string() }),
});
});
agent.serve();package main
import (
"context"
"fmt"
"log"
"github.com/Agent-Field/agentfield/sdk/go/agent"
"github.com/Agent-Field/agentfield/sdk/go/ai"
)
type Decision struct {
Action string `json:"action"`
}
func main() {
a, err := agent.New(agent.Config{
NodeID: "demo",
Version: "1.0.0",
AgentFieldURL: "http://localhost:8080",
AIConfig: &ai.Config{Model: "anthropic/claude-sonnet-4-20250514"},
})
if err != nil {
log.Fatal(err)
}
a.RegisterReasoner("route", func(ctx context.Context, input map[string]any) (any, error) {
text := fmt.Sprintf("%v", input["text"])
resp, err := a.AI(ctx, text,
ai.WithSystem("Pick one action: summarize | escalate | done."),
ai.WithSchema(Decision{}),
)
if err != nil {
return nil, err
}
var d Decision
resp.JSON(&d)
return d, nil
})
a.Serve(context.Background())
}- APIs — decorated functions become HTTP endpoints with discovery and tracing.
- Models — 100+ LLMs, structured output (Pydantic / Zod / structs), tool calling.
- Multi-agent —
app.call, shared memory, async, webhooks, governance (DIDs, policy, audit).
Quickstart
Build with your coding agent, or write the first agent yourself.
AgentField vs frameworks
When an agent becomes an AI backend, not just an app feature.
How AgentField works
Primitives, control plane, sync, async, and cross-agent calls.
Production capabilities
Agent APIs, triggers, IAM, rollouts, audit, and SDKs.
Build
Agents, intelligence, coordination, execution, governance
Reference
SDKs, REST, CLI, deploy, testing