Build powerful AI assistants by creating specialized agent teams that work together.
10K+
This guide explains how to build and run AgentCrew using Docker.
# Run with console interface
docker run -it --rm \
-e ANTHROPIC_API_KEY="your_claude_api_key" \
-e OPENAI_API_KEY="your_openai_api_key" \
daltonnyx/agentcrew chat --console
# Run as A2A server
docker run -d \
--name agentcrew-server \
-p 41241:41241 \
-e ANTHROPIC_API_KEY="your_claude_api_key" \
daltonnyx/agentcrew a2a-server --host 0.0.0.0 --port 41241
# Create a named volume for persistence
docker volume create agentcrew_data
# Run with persistent data
docker run -it --rm \
-v agentcrew_data:/home/agentcrew/.AgentCrew \
-e ANTHROPIC_API_KEY="your_api_key" \
daltonnyx/agentcrew chat --console
# Create local config directory
mkdir -p ~/.agentcrew-docker
# Run with host directory mounting
docker run -it --rm \
-v ~/.agentcrew-docker:/home/agentcrew/.AgentCrew \
-e ANTHROPIC_API_KEY="your_api_key" \
daltonnyx/agentcrew chat --console
Create a custom agents.toml file:
[[agents]]
name = "researcher"
description = "AI Research Assistant"
system_prompt = """You are a research assistant specialized in finding and analyzing information.
Current date: {current_date}
"""
tools = ["memory", "web_search", "code_analysis"]
[[agents]]
name = "coder"
description = "AI Coding Assistant"
system_prompt = """You are a coding assistant specialized in software development.
Current date: {current_date}
"""
tools = ["memory", "clipboard", "code_analysis"]
Mount it when running:
docker run -it --rm \
-v $(pwd)/agents.toml:/home/agentcrew/.AgentCrew/agents.toml:ro \
-e ANTHROPIC_API_KEY="your_api_key" \
daltonnyx/agentcrew chat --console --agent-config /home/agentcrew/.AgentCrew/agents.toml
Create a config.json file:
{
"api_keys": {
"ANTHROPIC_API_KEY": "your_claude_api_key",
"OPENAI_API_KEY": "your_openai_api_key",
"GROQ_API_KEY": "your_groq_api_key"
}
}
Mount it when running:
docker run -it --rm \
-v $(pwd)/config.json:/home/agentcrew/.AgentCrew/config.json:ro \
daltonnyx/agentcrew chat --console
# Console mode with specific provider
docker run -it --rm daltonnyx/agentcrew chat --console --provider openai
# With custom configurations
docker run -it --rm \
-v $(pwd)/custom_agents.toml:/home/agentcrew/.AgentCrew/agents.toml:ro \
daltonnyx/agentcrew chat --console --agent-config /home/agentcrew/.AgentCrew/agents.toml
# Basic server
docker run -d -p 41241:41241 daltonnyx/agentcrew a2a-server
# Server with specific provider and API key
docker run -d -p 41241:41241 \
-e OPENAI_API_KEY="your_key" \
daltonnyx/agentcrew a2a-server --provider openai --host 0.0.0.0 --port 41241
# Server with authentication
docker run -d -p 41241:41241 \
daltonnyx/agentcrew a2a-server --api-key "your_server_auth_key"
# Authenticate with GitHub Copilot (interactive)
docker run -it --rm \
-v agentcrew_data:/home/agentcrew/.AgentCrew \
daltonnyx/agentcrew copilot-auth
docker volume prune to clean up unused volumes# 1. Create a project directory
mkdir agentcrew-docker && cd agentcrew-docker
# 2. Create environment file
cat > .env << EOF
ANTHROPIC_API_KEY=your_claude_api_key
OPENAI_API_KEY=your_openai_api_key
GROQ_API_KEY=your_groq_api_key
EOF
# 3. Create custom agents configuration
cat > agents.toml << EOF
[[agents]]
name = "assistant"
description = "General AI Assistant"
system_prompt = """You are a helpful AI assistant.
Current date: {current_date}
"""
tools = ["memory", "clipboard", "web_search", "code_analysis"]
EOF
# 4. Build and run
docker run -it --rm \
--env-file .env \
-v $(pwd)/agents.toml:/home/agentcrew/.AgentCrew/agents.toml:ro \
-v agentcrew_data:/home/agentcrew/.AgentCrew \
daltonnyx/agentcrew chat --console
# Run as production A2A server with restart policy
docker run -d \
--name agentcrew-prod \
--restart unless-stopped \
-p 41241:41241 \
-v agentcrew_prod_data:/home/agentcrew/.AgentCrew \
-e ANTHROPIC_API_KEY="your_api_key" \
daltonnyx/agentcrew a2a-server \
--host 0.0.0.0 \
--port 41241 \
--api-key "your_server_auth_key"
API Keys: Never include API keys in the Docker image. Always use environment variables or mounted config files.
Authentication: Use the --api-key option for A2A server mode in
production.
Network: Consider using Docker networks or reverse proxies for production deployments.
File Permissions: The container runs as non-root user agentcrew for
security.
If you want to modify the image:
# Clone the repository
git clone https://github.com/saigontechnology/AgentCrew.git
cd AgentCrew
# Build custom image
docker build -t my-agentcrew . -f docker/Dockerfile
# Or with build arguments (if needed)
docker build --build-arg PYTHON_VERSION=3.12 -t my-agentcrew .
Content type
Image
Digest
sha256:d95a7ee40…
Size
1.1 GB
Last updated
8 days ago
docker pull daltonnyx/agentcrew