Bernstein Multi-Agent
🐍 🏠 ☁️ 🍎 🪟 🐧 - Deterministic multi-agent orchestrator for 37 CLI coding agents (Claude C
What is Bernstein Multi-Agent?
Bernstein Multi-Agent is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to 🐍 🏠 ☁️ 🍎 🪟 🐧 - deterministic multi-agent orchestrator for 37 cli coding agents (claude c
🐍 🏠 ☁️ 🍎 🪟 🐧 - Deterministic multi-agent orchestrator for 37 CLI coding agents (Claude C
This server falls under the Coding Agents category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- MCP protocol support
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx bernstein-multi-agent-orchestrationConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use Bernstein Multi-Agent
Bernstein is a deterministic multi-agent orchestration platform that coordinates up to 44 CLI coding agents — including Claude Code, Codex CLI, Aider, Gemini CLI, and GitHub Copilot CLI — running in parallel across isolated Git worktrees. Its Python scheduler contains zero LLM calls in the coordination loop, making orchestration fully deterministic, and it enforces per-artifact lineage tracking, HMAC-SHA256 audit chains, and per-session zero-trust JWT authentication. Engineering teams use Bernstein to parallelize large refactors, run cross-model code reviews, and automate the idea-to-PR pipeline with quality gates including tests, linting, and type checking.
Prerequisites
- Python 3.12 or later
- Git with worktree support (Git 2.5+)
- At least one CLI coding agent installed (Bernstein auto-discovers available agents on startup)
- ANTHROPIC_API_KEY for Claude-based agents, or the appropriate API key for your chosen agent
- An MCP-compatible client if using Bernstein's MCP server mode; Docker for container sandbox isolation
Install Bernstein
Install Bernstein using pip, pipx, or uv. The pipx method is recommended as it isolates Bernstein in its own environment.
pipx install bernsteinSet your API key
Export the API key for the LLM provider powering your coding agents. For Claude Code agents, set ANTHROPIC_API_KEY.
export ANTHROPIC_API_KEY=sk-ant-...Initialize the workspace
Run bernstein init inside your project repository. This creates the .sdd/ state directory and a bernstein.yaml configuration file.
cd /path/to/your/project
bernstein initRun your first orchestrated task
Pass a goal with the -g flag. Bernstein spawns agents in parallel Git worktrees, each working on the task independently, then merges and verifies the results.
bernstein -g "Add input validation to all API endpoints and write unit tests"Monitor progress in the TUI dashboard
In a separate terminal, open the live dashboard to watch agent progress, view per-agent diffs, and see quality gate results.
bernstein liveCreate a pull request from the session
Once agents complete and all quality gates pass, create a GitHub PR from the best-scoring worktree automatically.
bernstein prBernstein Multi-Agent Examples
Client configuration
Claude Desktop configuration to connect to Bernstein running as an MCP server, with the Anthropic API key for Claude-based agents.
{
"mcpServers": {
"bernstein": {
"command": "bernstein",
"args": ["mcp", "serve"],
"env": {
"ANTHROPIC_API_KEY": "sk-ant-..."
}
}
}
}Prompts to try
Example tasks to orchestrate via Bernstein from an MCP client or the CLI.
- "Use Bernstein to parallelize refactoring the authentication module across 3 Claude Code agents"
- "Run a cross-model code review using both Claude Code and Gemini CLI on the payment service"
- "Set up a recurring schedule to run the test suite and auto-fix failures every night"
- "Start a Bernstein workflow to take the idea of adding JWT auth from spec to a merged PR"Troubleshooting Bernstein Multi-Agent
Bernstein cannot find any coding agents and exits immediately
Bernstein auto-discovers agents by checking for executables on the PATH. Install at least one agent first — for Claude Code run `npm install -g @anthropic-ai/claude-code`, then verify discovery with `bernstein agents list`.
Git worktree creation fails with 'fatal: not a git repository'
Bernstein must be run from inside a Git repository. Navigate to your project root and run `git init` if no repository exists, then `bernstein init` to create the workspace state.
Agents complete but quality gates (tests/lint) fail and the session does not produce a PR
Quality gates are configured in bernstein.yaml. Check the gate configuration and ensure the test/lint commands are correct for your project. You can run `bernstein autofix` to spawn a fixer agent that addresses gate failures automatically.
Frequently Asked Questions about Bernstein Multi-Agent
What is Bernstein Multi-Agent?
Bernstein Multi-Agent is a Model Context Protocol (MCP) server that 🐍 🏠 ☁️ 🍎 🪟 🐧 - deterministic multi-agent orchestrator for 37 cli coding agents (claude c It connects AI assistants to external tools and data sources through a standardized interface.
How do I install Bernstein Multi-Agent?
Follow the installation instructions on the Bernstein Multi-Agent GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with Bernstein Multi-Agent?
Bernstein Multi-Agent works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is Bernstein Multi-Agent free to use?
Yes, Bernstein Multi-Agent is open source and available under the Apache 2.0 license. You can use it freely in both personal and commercial projects.
Bernstein Multi-Agent Alternatives — Similar Coding Agents Servers
Looking for alternatives to Bernstein Multi-Agent? Here are other popular coding agents servers you can use with Claude, Cursor, and VS Code.
Dify
★ 142.2kProduction-ready platform for agentic workflow development.
Ruflo
★ 54.0k🌊 The leading agent orchestration platform for Claude. Deploy intelligent multi-agent swarms, coordinate autonomous workflows, and build conversational AI systems. Features enterprise-grade architecture, self-learning swarm intelligence, RAG integrat
Goose
★ 45.7kan open source, extensible AI agent that goes beyond code suggestions - install, execute, edit, and test with any LLM
Antigravity Awesome Skills
★ 38.3kInstallable GitHub library of 1,400+ agentic skills for Claude Code, Cursor, Codex CLI, Gemini CLI, Antigravity, and more. Includes installer CLI, bundles, workflows, and official/community skill collections.
AgentScope
★ 25.5kBuild and run agents you can see, understand and trust.
Serena
★ 24.5kA coding agent toolkit that provides IDE-like semantic code retrieval and editing tools, enabling LLMs to efficiently navigate and modify codebases using symbol-level operations instead of basic file reading and string replacements.
Browse More Coding Agents MCP Servers
Explore all coding agents servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.
Set Up Bernstein Multi-Agent in Your Editor
Choose your AI client for step-by-step setup instructions.
Quick Config Preview
Add this to your claude_desktop_config.json or .cursor/mcp.json
Ready to use Bernstein Multi-Agent?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.