Skip to content
Control Center

Manage adapters and models

This guide shows you how to manage adapters and models.

What are adapters?

Section titled “What are adapters?”

An adapter is a CLI backend that Control Center uses to dispatch agents. Each adapter wraps a specific AI CLI tool:

AdapterCLI nameDescription
Claude CodeclaudeAnthropic’s Claude CLI
CodexcodexOpenAI’s Codex CLI
PipiPi agent runner

View detected adapters

Section titled “View detected adapters”
  1. Navigate to Settings → Adapters
  2. The list shows all detected adapters with:
    • Detection status (found/not found)
    • CLI version
    • Installation path

Control Center auto-detects CLIs in your PATH on startup.

Install an adapter

Section titled “Install an adapter”

Install the CLI using its package manager:

Terminal window
# Claude Code
npm install -g @anthropic-ai/claude-code


# Codex
npm install -g @openai/codex

After installation, refresh the adapter list in Settings.

Select a model

Section titled “Select a model”

Each adapter offers multiple models. When configuring an agent:

  1. Select the adapter
  2. The model dropdown shows available models for that adapter
  3. Pick the model based on your needs:
Model tierUse for
Fast/cheapSimple tasks, formatting, quick fixes
Mid-tierTypical coding, bug fixes
Top-tierComplex architecture, critical code, security review

Model availability depends on your adapter and API key.

Custom adapters

Section titled “Custom adapters”

Register custom adapters in Settings:

  1. Click Add adapter
  2. Enter the CLI name and path
  3. Control Center probes the CLI for version and capabilities
  4. The adapter appears in the agent configuration dropdown

ACP models

Section titled “ACP models”

Control Center also supports ACP-compatible agent runners. Advertised models are keyed by adapter ID and shown in the model selector.

Troubleshooting

Section titled “Troubleshooting”

Adapter not detected

Section titled “Adapter not detected”
  • Ensure the CLI is in your PATH (which claude)
  • Restart Control Center after installing a new CLI
  • Check the adapter settings for the expected CLI name

Model not available

Section titled “Model not available”
  • Verify your API key has access to the model
  • Check the adapter’s documentation for supported models
  • Some models require specific CLI versions
Section titled “Related guides”