Code Genesis#

Code Genesis is an open-source multi-agent framework that generates production-ready software projects from natural language requirements. It orchestrates specialized AI agents to autonomously deliver end-to-end project generation with frontend, backend, and database integration.

Overview#

Features#

End-to-end project generation: Automatically generates complete projects with frontend, backend, and database integration from natural language descriptions
High-quality code: LSP validation and dependency resolution ensure production-ready output
Topology-aware generation: Eliminates reference errors through dependency-driven code generation
Automated deployment: Deploys to EdgeOne Pages automatically with MCP integration
Flexible workflows: Choose between standard (7-stage) or simple (4-stage) pipelines based on project complexity

Architecture#

Code Genesis provides two configurable workflow modes:

Standard Workflow (Production-Grade)#

Standard Workflow

The standard pipeline implements a rigorous 7-stage process optimized for complex, production-ready projects:

User Story → Architect → File Design → File Order → Install → Coding → Refine

Pipeline Stages:

User Story Agent: Parses user requirements into structured user stories
Architect Agent: Selects technology stack and defines system architecture
File Design Agent: Generates physical file structure from architectural blueprint
File Order Agent: Constructs dependency DAG and topological sort for parallel code generation
Install Agent: Bootstraps environment and resolves dependencies
Coding Agent: Synthesizes code with LSP validation, following dependency order
Refine Agent: Performs runtime validation, bug fixing, and automated deployment

Each agent produces structured intermediate outputs, ensuring engineering rigor throughout the pipeline.

Simple Workflow (Rapid Prototyping)#

Simple Workflow

For lightweight projects or quick iterations, the simple workflow condenses the pipeline into 4 core stages:

Orchestrator → Install → Coding → Refine

Streamlined Process:

Orchestrator Agent: Unified requirement analysis, architecture design, and file planning
Install Agent: Dependency resolution and environment setup
Coding Agent: Direct code generation with integrated file ordering
Refine Agent: Validation and deployment

Workflow Comparison#

Aspect	Standard Workflow	Simple Workflow
Agent Stages	7 specialized agents	4 consolidated agents
Architecture Quality	Explicit, auditable design	Implicit, monolithic design
Generation Time	Moderate (thorough planning)	Fast (direct execution)
Use Cases	Production systems, complex apps	Prototypes, demos, simple tools

How to Use#

Installation#

Clone the repository and prepare the environment:

git clone https://github.com/modelscope/ms-agent
cd ms-agent
pip install -r requirements/code.txt
pip install -e .

Prepare npm environment, following https://nodejs.org/en/download. If you are using Mac, using Homebrew is recommended: https://formulae.brew.sh/formula/node

Make sure your installation is successful:

npm --version

Make sure the npm installation is successful, or the npm install/build/dev will fail.

Quick Start#

Run the standard workflow:

PYTHONPATH=. openai_api_key=your-api-key openai_base_url=your-api-url \
python ms_agent/cli/cli.py run \
  --config projects/code_genesis \
  --query 'make a demo website' \
  --trust_remote_code true

The code will be output to the output folder in the current directory by default.

Advanced Configuration#

Enable Diff-Based File Editing#

Add edit_file_config to both coding.yaml and refine.yaml:

edit_file_config:
  model: morph-v3-fast  # or other compatible models
  api_key: your-api-key
  base_url: https://api.morphllm.com/v1

Get your model and API key from https://www.morphllm.com

Enable Automated Deployment#

Add edgeone-pages-mcp configuration to refine.yaml:

mcp_servers:
  edgeone-pages:
    env:
      EDGEONE_PAGES_API_TOKEN: your-edgeone-token

Get your EDGEONE_PAGES_API_TOKEN from https://pages.edgeone.ai/zh/document/pages-mcp