CONDUCTOR.md

Conductor Setup

Automatic Setup

When you create a new Conductor worktree, the environment is automatically configured via .conductor.json.

The setup script copies .env from the repository root to your worktree, giving you instant access to configured agents.

First-Time Setup

1. Configure Repository Root

In the repository root (not in a worktree):

cd /path/to/adcp-client-python
cp .env.example .env
# Edit .env with your agent configurations

2. Start Working

That's it! When you create or enter a Conductor worktree, the setup happens automatically.

Environment Variables

ADCP_AGENTS (Recommended)

Use this JSON format for multi-agent configuration:

ADCP_AGENTS='[
  {
    "id": "creative_agent",
    "name": "Creative Agent",
    "agent_uri": "https://creative.adcontextprotocol.org",
    "protocol": "mcp"
  },
  {
    "id": "my_agent",
    "name": "My Agent",
    "agent_uri": "https://my-agent.example.com",
    "protocol": "mcp",
    "auth_token": "your-token-here"
  }
]'

Then in code:

from adcp import ADCPMultiAgentClient

client = ADCPMultiAgentClient.from_env()
agent = client.agent("my_agent")

Individual Variables

For single-agent usage:

AGENT_URI=https://agent.example.com
AGENT_PROTOCOL=mcp
AGENT_TOKEN=your-token

Manual Setup

If automatic setup doesn't work, run manually:

python3 scripts/setup_conductor_env.py

Security

• ✅ .env is gitignored
• ✅ Store tokens in .env at repository root
• ✅ Use .env.example for documentation
• ❌ Never commit actual tokens to git

Testing

# Creative agent (public, no auth)
python3 tests/integration/test_creative_agent.py

# Your own agents (configured in .env)
python3 tests/integration/test_your_agent.py

Troubleshooting

"No .env file found":

cd /path/to/adcp-client-python  # Repository root
cp .env.example .env
# Edit .env

Setup didn't run automatically:

# Run manually in your worktree
python3 scripts/setup_conductor_env.py

See full documentation in README.md for complete usage examples.