@funtuantw/pi-agent-cf

Deploy pi-mono AI agents on Cloudflare Workers with Durable Objects.

Dependency-injection friendly SDK — inject your own tools, LLM providers, API keys, and storage in a few lines.

Features

• WebSocket streaming — real-time bidirectional communication via Durable Objects Hibernation API
• REST API — create sessions, query state, send prompts over HTTP
• Session persistence — messages auto-saved to DO storage, survives restarts
• Dependency injection — factory function accepts tools, API keys, system prompt, event hooks, auth middleware
• Multi-client — multiple WebSocket clients can share the same agent session
• Idle cleanup — configurable alarm auto-evicts unused sessions from memory

Quick Start

1. Install

npm install @funtuantw/pi-agent-cf @mariozechner/pi-agent-core @mariozechner/pi-ai

2. Create your worker

// src/index.ts
import { createAgentWorker, type AgentEnv, type AgentTool } from '@funtuantw/pi-agent-cf';
import { Type } from '@sinclair/typebox';

interface Env extends AgentEnv {
  GOOGLE_API_KEY?: string;
}

// Define a custom tool
const GetTime: AgentTool<typeof Type.Object({})> = {
  name: 'get_time',
  label: 'Get Time',
  description: 'Returns the current UTC time.',
  parameters: Type.Object({}),
  execute: async () => ({
    content: [{ type: 'text', text: new Date().toISOString() }],
    details: null,
  }),
};

// Create the worker with DI config
const worker = createAgentWorker<Env>({
  systemPrompt: 'You are a helpful assistant.',
  tools: () => [GetTime],
  getApiKey: (provider, env) => {
    if (provider === 'google') return env.GOOGLE_API_KEY;
  },
});

export const AgentSessionDO = worker.AgentSessionDO;
export default worker.handler;

3. Configure wrangler.toml

name = "my-agent"
main = "src/index.ts"
compatibility_date = "2025-01-01"
compatibility_flags = ["nodejs_compat"]

[durable_objects]
bindings = [
  { name = "AGENT_SESSION", class_name = "AgentSessionDO" }
]

[[migrations]]
tag = "v1"
new_classes = ["AgentSessionDO"]

# Required: AJV uses new Function() which is forbidden in CF Workers.
# This stub disables schema validation; pi-ai falls back to trusting LLM output.
[alias]
"ajv" = "node_modules/@funtuantw/pi-agent-cf/stubs/ajv.js"

4. Set secrets & deploy

wrangler secret put GOOGLE_API_KEY
wrangler deploy

API

REST Endpoints

WebSocket Protocol

Connect to /sessions/:id/ws and send JSON messages:

Client → Server:

{ "type": "prompt", "text": "Hello!" }           // Start a conversation
{ "type": "steer", "message": { ... } }           // Inject mid-turn
{ "type": "follow_up", "message": { ... } }       // Queue post-turn
{ "type": "abort" }                                // Stop generation
{ "type": "get_state" }                            // Query state
{ "type": "set_model", "provider": "anthropic", "modelId": "claude-sonnet-4-20250514" }
{ "type": "set_thinking_level", "level": "medium" }
{ "type": "clear_messages" }                       // Clear history
{ "type": "reset" }                                // Full reset
{ "type": "restore" }                              // Restore from storage
{ "type": "ping" }

Server → Client:

{ "type": "event", "event": { "type": "message_update", ... } }  // AgentEvent stream
{ "type": "state", "state": { ... } }              // State response
{ "type": "restored", "messages": [...] }           // Restored messages
{ "type": "error", "message": "...", "code": "..." }
{ "type": "session_created", "sessionId": "..." }
{ "type": "pong" }

Configuration (AgentWorkerConfig)

Architecture

Client (Browser/App)
    ↕ WebSocket / REST
CF Worker (Router)
    ↕ Durable Object stub
AgentSession DO (one per session)
    ├── Agent instance (pi-agent-core)
    ├── StreamFn → LLM Provider (pi-ai fetch)
    ├── Custom Tools (injected via config)
    └── DO Storage (persistent messages)

• Worker Router — handles HTTP routing, CORS, auth, creates DO stubs
• AgentSession DO — one instance per conversation session
  • Manages the Agent lifecycle from pi-agent-core
  • Accepts WebSocket connections with Hibernation API (cost-efficient)
  • Broadcasts all AgentEvents to connected clients
  • Persists messages to DO storage on each turn completion
  • Cleans up in-memory resources after idle timeout via alarm

Compatibility

Important: pi-ai depends on AJV for JSON-schema validation, which uses new Function() internally — forbidden in Cloudflare Workers. The SDK ships a stub at stubs/ajv.js that disables AJV so pi-ai falls back to trusting the LLM output directly. You must add the [alias] section to your wrangler.toml (see step 3 above).

Also, if you use undici >= 7 (pulled by pi-ai), add "undici": "^6.21.0" to your overrides in package.json to avoid node:sqlite bundling issues.

The SDK uses only pi-agent-core (zero Node.js deps) and pi-ai (uses fetch() for LLM calls). The following providers work out of the box on CF Workers:

• ✅ Google (Gemini)
• ✅ Anthropic (Claude)
• ✅ OpenAI (GPT, o-series)
• ✅ Azure OpenAI
• ✅ OpenAI-compatible (Groq, OpenRouter, etc.)
• ❌ AWS Bedrock (requires AWS SDK)
• ❌ OAuth flows (require HTTP callback server)

License

MIT