Get started in 5 minutes ​

① Install the CLI ​

npm install -g agentproc

pipx install agentproc

pip install agentproc

Verify it works:

agentproc --version
# agentproc 0.2.0 (protocol 0.1)

② Pick a profile ​

Browse the Profile Hub — every profile is a directory containing profile.yaml, a bridge script, and a README. Five official profiles to start with:

③ Run it ​

Clone the repo and try echo-agent first (no API key needed):

git clone https://github.com/jeffkit/agentproc
cd agentproc

agentproc --profile hub/echo-agent/profile.yaml \
          --prompt "hello" \
          --cwd hub/echo-agent
# → You said: hello

Now a real one. With claude-code, you get streaming output and multi-turn session continuity:

agentproc --profile hub/claude-code/profile.yaml \
          --prompt "what is this codebase?" \
          --cwd ~/projects/my-app \
          --env ANTHROPIC_API_KEY="$ANTHROPIC_API_KEY"

You'll see protocol lines stream on stderr in real time, and the final reply on stdout:

AGENT_PARTIAL:"This codebase is..."
AGENT_SESSION:13c2f6ec-1f97-42c4-be9e-9475129e243c
agentproc:session:13c2f6ec-1f97-42c4-be9e-9475129e243c

Capture that session id and continue the conversation:

agentproc --profile hub/claude-code/profile.yaml \
          --prompt "tell me about the auth module" \
          --cwd ~/projects/my-app \
          --env ANTHROPIC_API_KEY="$ANTHROPIC_API_KEY" \
          --session 13c2f6ec-1f97-42c4-be9e-9475129e243c

④ Connect to your messaging platform ​

AgentProc agents don't talk to WeChat or Slack directly — that's the bridge's job. The bridge is a small program that:

1. Receives a message from the messaging platform (via webhook, polling, etc.)
2. Spawns your agent with AGENT_MESSAGE env var set
3. Reads the agent's stdout (per the AgentProc protocol)
4. Forwards the reply back to the user

Here's a complete working bridge in ~30 lines of Node.js that wires agentproc to anything:

// bridge.js — a minimal AgentProc bridge
const { run } = require('agentproc');
const fs = require('fs');

async function handleMessage(message, sessionId) {
  const profile = JSON.parse(fs.readFileSync('./profile.json'));
  const result = await run(profile, {
    message,
    sessionId,
    onPartial: (chunk) => console.log(`[streaming] ${chunk}`),
  });
  console.log(`Reply: ${result.reply}`);
  console.log(`Session: ${result.sessionId}`);  // pass this back next turn
}

// Replace with your platform's SDK:
// yourMessagingPlatform.onMessage(handleMessage);
handleMessage(process.argv[2] || 'hello', '');

Save as bridge.js, point it at a profile, and wire it to your messaging platform's webhook. The runner.js source is the spec in code form — read it as the canonical reference.

Where to go next ​

• Read the protocol spec — 1 page, defines everything
• Profile Hub — drop-in profiles for popular CLIs
• CLI reference — every flag and option
• Python SDK / Node SDK — embed AgentProc in your bridge
• Examples — claude_code bridge, bare scripts, more