Skip to content

Installation

CodeTether Server can be installed in several ways depending on your needs.

Requirements

  • Python 3.10+ (for pip install)
  • Docker 20.10+ (for container deployment)
  • Redis 6+ (optional, for distributed workers)

Install via pip (Agent Worker)

The pip package installs the Agent Worker - the component that executes tasks on machines where your code lives. The server itself is typically deployed via Docker or Kubernetes.

What's in the pip package?

  • codetether-worker — Agent worker CLI that polls for tasks and runs CodeTether
  • codetether — Server CLI (for local development/testing)
  • a2a-server — Alias for server CLI
pip install codetether

Verify the installation:

codetether-worker --version
# 1.4.3

Install from GitHub (pip)

If you want the latest code (or you're installing before a PyPI release is published):

pip install "git+https://github.com/rileyseaburg/codetether.git@main"

One-line install (shell)

This installs CodeTether into a local virtual environment under ~/.codetether/venv and links the CLI into ~/.local/bin.

Linux/macOS (Bash):

curl -fsSL https://raw.githubusercontent.com/rileyseaburg/codetether/main/scripts/install-codetether.sh | bash

Windows (PowerShell):

Invoke-Expression (Invoke-WebRequest -Uri "https://raw.githubusercontent.com/rileyseaburg/A2A-Server-MCP/main/scripts/install-codetether.ps1" -UseBasicParsing).Content

Docker

Pull the official image:

docker pull ghcr.io/rileyseaburg/codetether-server:latest

Run the server:

docker run -d \
  --name codetether \
  -p 8000:8000 \
  -p 9000:9000 \
  ghcr.io/rileyseaburg/codetether-server:latest

With environment configuration:

docker run -d \
  --name codetether \
  -p 8000:8000 \
  -p 9000:9000 \
  -e A2A_AGENT_NAME="My Agent" \
  -e REDIS_URL="redis://redis:6379" \
  -e KEYCLOAK_URL="https://auth.example.com" \
  ghcr.io/rileyseaburg/codetether-server:latest

Docker Compose

Create a docker-compose.yml:

version: '3.8'

services:
  codetether:
    image: ghcr.io/rileyseaburg/codetether-server:latest
    ports:
      - "8000:8000"  # A2A API
      - "9000:9000"  # MCP Server
    environment:
      - A2A_AGENT_NAME=CodeTether Server
      - REDIS_URL=redis://redis:6379
      - LOG_LEVEL=INFO
    depends_on:
      - redis

  redis:
    image: redis:7-alpine
    ports:
      - "6379:6379"
    volumes:
      - redis-data:/data

volumes:
  redis-data:

Start the stack:

docker-compose up -d

From Source

Clone the repository:

git clone https://github.com/rileyseaburg/codetether.git
cd codetether

Create a virtual environment:

python -m venv venv
source venv/bin/activate  # Linux/macOS
# or
.\venv\Scripts\activate   # Windows

Install dependencies:

pip install -e .

Run the server:

codetether --port 8000

CodeTether (Local Fork)

CodeTether includes a maintained fork of CodeTether in the agent/ directory. To build it:

cd agent

# Install dependencies
bun install

# Build the project
bun run build

# Verify installation
agent --version

For more details, see CodeTether Integration.

Agent Worker (systemd, Linux)

The agent worker executes CodeTether tasks on machines where your code lives. It polls the server for tasks and runs them in the configured codebase directories.

Prerequisites

  1. Build CodeTether from the local fork
  2. Python 3.10+ with a virtual environment

Installation

# Clone and set up
git clone https://github.com/rileyseaburg/codetether.git
cd codetether

# Create virtual environment
python -m venv .venv
source .venv/bin/activate

# Install the worker
pip install codetether

# Run the installer (sets up systemd service)
sudo ./agent_worker/install.sh

# Start the worker
sudo systemctl start a2a-agent-worker
sudo journalctl -u a2a-agent-worker -f

Upgrading the Worker

To upgrade to a new version:

# Activate the virtual environment used by the worker
source /path/to/your/venv/bin/activate

# Upgrade codetether
pip install --upgrade codetether

# Restart the service
sudo systemctl restart a2a-agent-worker

Configuration

Edit /etc/a2a-agent-worker/config.json to configure:

  • server_url — URL of the CodeTether server
  • codebases — List of directories to register
  • agent_name — Name for this worker
  • models_supported — LLM models this worker can use

Kubernetes (Helm)

Add the Helm repository:

helm repo add codetether https://charts.codetether.run
helm repo update

Install the chart:

helm install codetether codetether/a2a-server \
  --namespace codetether \
  --create-namespace \
  --set ingress.enabled=true \
  --set ingress.hosts[0].host=codetether.example.com

See Kubernetes Deployment for full configuration options.

Verify Installation

Once installed, verify the server is running:

# Check health endpoint
curl http://localhost:8000/health
# {"status": "healthy", "version": "1.0.0"}

# Check agent card (A2A discovery)
curl http://localhost:8000/.well-known/agent-card.json

Next Steps