Bamboo Agent ๐
A fully self-contained AI agent backend framework with built-in web services.
Bamboo provides everything you need to build, deploy, and scale AI-powered applications with support for multiple LLM providers, comprehensive tool execution, and production-ready infrastructure.
โจ Features
- ๐ค Complete Agent System: Agent loop, tool execution, skill management
- ๐ Built-in HTTP Server: Actix-web based API server with REST and streaming endpoints
- ๐ง Multi-LLM Support: OpenAI, Anthropic, Google Gemini, GitHub Copilot
- ๐ XDG-Compliant: Follows XDG Base Directory specification
- ๐ง Dual Mode: Binary (standalone server) or library (embedded)
- ๐ Production-Ready: CORS, rate limiting, security headers built-in
- ๐ Session Management: Persistent conversation history
- โก Workflow System: Automate complex tasks with YAML workflows
- ๐ค Claude Integration: Seamless Claude Code binary discovery and management
- ๐งช Well Tested: 806 tests with 100% pass rate
Installation
From crates.io
From source
Quick Start
Binary Mode
# Start server with default settings
# Custom configuration
Library Mode
use ;
async
Configuration
Bamboo follows the XDG Base Directory specification.
Default Paths
- Config:
$XDG_CONFIG_HOME/bamboo/config.json(default:~/.config/bamboo/) - Data:
$XDG_DATA_HOME/bamboo/(default:~/.local/share/bamboo/) - Cache:
$XDG_CACHE_HOME/bamboo/(default:~/.cache/bamboo/)
Configuration File
Edit ~/.config/bamboo/config.json:
Environment Variables
Override configuration with environment variables:
BAMBOO_PORT: Server portBAMBOO_BIND: Bind addressBAMBOO_DATA_DIR: Data directory
API Endpoints
Once running, Bamboo exposes the following endpoints:
Health Check
GET /api/v1/health
Chat Completions
POST /api/v1/chat/completions
Agent Execution
POST /api/v1/agent/run
Workflows
GET /v1/workflows
POST /v1/workflows
DELETE /v1/workflows/{name}
Sessions
GET /api/v1/sessions
POST /api/v1/sessions
Development
Build
Test
Run
Architecture
Bamboo is organized into the following modules:
config: Configuration management with XDG supportcore: Core types and utilitiesagent: Agent system (loop, tools, skills, LLM providers)server: HTTP server and controllersprocess: Process managementclaude: Claude Code integrationcommands: Workflow, slash commands, keyword masking
License
MIT
Contributing
Contributions are welcome! Please read our contributing guidelines before submitting PRs.
Roadmap
- Complete agent system migration
- Full OpenAI/Anthropic API compatibility
- Webhook support
- Plugin system
- gRPC API
- Kubernetes deployment guides