CodeBoarding is an open-source codebase analysis tool that generates high-level diagram representations of codebases
using static analysis and LLM agents, that humans and agents can interact with.
It’s designed to support onboarding, documentation, and comprehension for large, complex systems.
- Extract modules and their relationships based on the control flow graph of the project.
- Builds different levels of abstraction with an LLM agent (multi-provider support) using remote or local inference.
- Outputs interactive diagrams (Mermaid.js) for integration into docs, IDEs, CI/CD.
📄 Existing visual generations: GeneratedOnBoardings
🌐 Try for your open-source project: www.codeboarding.org/diagrams
For detailed architecture information, see our diagram documentation.
graph LR
API_Service["API Service"]
Job_Database["Job Database"]
Orchestration_Engine["Orchestration Engine"]
Repository_Manager["Repository Manager"]
Static_Analysis_Engine["Static Analysis Engine"]
AI_Interpretation_Layer["AI Interpretation Layer"]
Output_Generation_Engine["Output Generation Engine"]
Unclassified["Unclassified"]
API_Service -- " Initiates Job " --> Job_Database
API_Service -- " Triggers Analysis " --> Orchestration_Engine
Orchestration_Engine -- " Manages Job State " --> Job_Database
Orchestration_Engine -- " Requests Code " --> Repository_Manager
Repository_Manager -- " Provides Code " --> Orchestration_Engine
Orchestration_Engine -- " Requests Static Analysis " --> Static_Analysis_Engine
Static_Analysis_Engine -- " Provides Richer Analysis Results " --> Orchestration_Engine
Orchestration_Engine -- " Feeds Rich Analysis Data " --> AI_Interpretation_Layer
AI_Interpretation_Layer -- " Returns Enhanced Architectural Insights " --> Orchestration_Engine
AI_Interpretation_Layer -- " Queries Diff " --> Repository_Manager
Orchestration_Engine -- " Passes Enhanced Insights for Generation " --> Output_Generation_Engine
Output_Generation_Engine -- " Delivers Documentation " --> API_Service
click Job_Database href "https://github.com/CodeBoarding/CodeBoarding/blob/main/.codeboarding/Job_Database.md" "Details"
click Orchestration_Engine href "https://github.com/CodeBoarding/CodeBoarding/blob/main/.codeboarding/Orchestration_Engine.md" "Details"
click Repository_Manager href "https://github.com/CodeBoarding/CodeBoarding/blob/main/.codeboarding/Repository_Manager.md" "Details"
click Static_Analysis_Engine href "https://github.com/CodeBoarding/CodeBoarding/blob/main/.codeboarding/Static_Analysis_Engine.md" "Details"
click AI_Interpretation_Layer href "https://github.com/CodeBoarding/CodeBoarding/blob/main/.codeboarding/AI_Interpretation_Layer.md" "Details"
click Output_Generation_Engine href "https://github.com/CodeBoarding/CodeBoarding/blob/main/.codeboarding/Output_Generation_Engine.md" "Details"
First, make you you have uv installed. Check official installation guide Installing UV.
Setup the environment:
uv sync --frozen
source .venv/bin/activate
python setup.pyImportant
The setup script installs a language server for Python and TypeScript/JavaScript. In order to successfully install
the TypeScript Language Server, you need to have npm installed. If npm is not found, the script will skip the
installation of the TypeScript Language Server and you will need to install it manually later if you want to analyze
TypeScript/JavaScript projects.
Configure the environment variables in a .env file (you can copy from .env.example
8000
):
The python setup.py command creates a .env file if it doesn't exist with a default value for REPO_ROOT and ROOT_RESULT as well as OLLAMA_BASE_URL for local LLM inference. If you want to use a different LLM provider, you need to set the corresponding API key in the .env file.
# LLM Provider (choose one)
OPENAI_API_KEY=
ANTHROPIC_API_KEY=
GOOGLE_API_KEY=
AWS_BEARER_TOKEN_BEDROCK=
OLLAMA_BASE_URL=
OPENAI_BASE_URL= # Optional: Custom OpenAI endpoint
# Core Configuration
CACHING_DOCUMENTATION=false # Enable/disable documentation caching
REPO_ROOT=./repos # Directory for downloaded repositories
ROOT_RESULT=./results # Directory for generated outputs
PROJECT_ROOT=/path/to/CodeBoarding # Source project root (must end with /CodeBoarding)
DIAGRAM_DEPTH_LEVEL=1 # Max depth level for diagram generation
STATIC_ANALYSIS_CONFIG=./static_analysis_config.yml # Path to static analysis config
# Optional
GITHUB_TOKEN= # For accessing private repositories
LANGSMITH_TRACING=false # Optional: Enable LangSmith tracing
LANGSMITH_ENDPOINT= # Optional: LangSmith endpoint
LANGSMITH_PROJECT= # Optional: LangSmith project name
LANGCHAIN_API_KEY= # Optional: LangChain API key💡 Tip: Our experience has shown that using Google Gemini‑2.5‑Pro yields the best results for complex diagram generation tasks.
python demo.py <github_repo_url> --output-dir <output_path>We have visualized over 800+ popular open-source projects. See examples:
graph LR
Tensor_Operations_Core["Tensor Operations & Core"]
Automatic_Differentiation_Autograd_Engine_["Automatic Differentiation (Autograd Engine)"]
Neural_Network_Modules_torch_nn_["Neural Network Modules (torch.nn)"]
Optimizers_torch_optim_["Optimizers (torch.optim)"]
Data_Utilities_torch_utils_data_["Data Utilities (torch.utils.data)"]
JIT_Compiler_Scripting_TorchScript_["JIT Compiler & Scripting (TorchScript)"]
Hardware_Backends["Hardware Backends"]
Data_Utilities_torch_utils_data_ -- " provides data to " --> Tensor_Operations_Core
Tensor_Operations_Core -- " provides primitives for " --> Neural_Network_Modules_torch_nn_
Tensor_Operations_Core -- " leverages " --> Hardware_Backends
Neural_Network_Modules_torch_nn_ -- " performs operations on " --> Tensor_Operations_Core
Neural_Network_Modules_torch_nn_ -- " operations recorded by " --> Automatic_Differentiation_Autograd_Engine_
Neural_Network_Modules_torch_nn_ -- " exported to " --> JIT_Compiler_Scripting_TorchScript_
Automatic_Differentiation_Autograd_Engine_ -- " computes gradients for " --> Optimizers_torch_optim_
Optimizers_torch_optim_ -- " updates parameters of " --> Neural_Network_Modules_torch_nn_
Hardware_Backends -- " executes computations for " --> Tensor_Operations_Core
click Tensor_Operations_Core href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/pytorch/Tensor_Operations_Core.md" "Details"
click Automatic_Differentiation_Autograd_Engine_ href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/pytorch/Automatic_Differentiation_Autograd_Engine_.md" "Details"
click Neural_Network_Modules_torch_nn_ href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/pytorch/Neural_Network_Modules_torch_nn_.md" "Details"
click Optimizers_torch_optim_ href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/pytorch/Optimizers_torch_optim_.md" "Details"
click Data_Utilities_torch_utils_data_ href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/pytorch/Data_Utilities_torch_utils_data_.md" "Details"
click JIT_Compiler_Scripting_TorchScript_ href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/pytorch/JIT_Compiler_Scripting_TorchScript_.md" "Details"
click Hardware_Backends href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/pytorch/Hardware_Backends.md" "Details"
graph LR
Application_Core["Application Core"]
Middleware["Middleware"]
Routing["Routing"]
Request_Handling_Validation["Request Handling & Validation"]
Dependency_Injection["Dependency Injection"]
Security["Security"]
Response_Handling["Response Handling"]
API_Documentation["API Documentation"]
Application_Core -- " sends request to " --> Middleware
Middleware -- " forwards request to " --> Routing
Routing -- " uses " --> Request_Handling_Validation
Routing -- " uses " --> Dependency_Injection
Routing -- " provides data for " --> Response_Handling
Dependency_Injection -- " enables " --> Security
Response_Handling -- " sends response to " --> Middleware
API_Documentation -- " inspects " --> Routing
API_Documentation -- " inspects " --> Request_Handling_Validation
click Application_Core href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/fastapi/Application_Core.md" "Details"
click Middleware href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/fastapi/Middleware.md" "Details"
click Routing href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/fastapi/Routing.md" "Details"
click Request_Handling_Validation href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/fastapi/Request_Handling_Validation.md" "Details"
click Dependency_Injection href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/fastapi/Dependency_Injection.md" "Details"
click Security href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/fastapi/Security.md" "Details"
click API_Documentation href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/fastapi/API_Documentation.md" "Details"
graph LR
ChatTTS_Core_Orchestrator["ChatTTS Core Orchestrator"]
Text_Processing_Module["Text Processing Module"]
Speech_Synthesis_Models["Speech Synthesis Models"]
Velocity_Inference_Engine["Velocity Inference Engine"]
System_Utilities_Configuration["System Utilities & Configuration"]
ChatTTS_Core_Orchestrator -- " Orchestrates Text Flow " --> Text_Processing_Module
ChatTTS_Core_Orchestrator -- " Receives Processed Text " --> Text_Processing_Module
ChatTTS_Core_Orchestrator -- " Orchestrates Synthesis Flow " --> Speech_Synthesis_Models
ChatTTS_Core_Orchestrator -- " Receives Audio Output " --> Speech_Synthesis_Models
ChatTTS_Core_Orchestrator -- " Initializes & Configures " --> System_Utilities_Configuration
ChatTTS_Core_Orchestrator -- " Loads Assets " --> System_Utilities_Configuration
Text_Processing_Module -- " Receives Raw Text " --> ChatTTS_Core_Orchestrator
Text_Processing_Module -- " Provides Processed Text " --> ChatTTS_Core_Orchestrator
Speech_Synthesis_Models -- " Receives Processed Data " --> ChatTTS_Core_Orchestrator
Speech_Synthesis_Models -- " Generates Audio Output " --> ChatTTS_Core_Orchestrator
Speech_Synthesis_Models -- " Delegates Inference To " --> Velocity_Inference_Engine
Speech_Synthesis_Models -- " Receives Inference Results " --> Velocity_Inference_Engine
Speech_Synthesis_Models -- " Utilizes GPU Resources " --> System_Utilities_Configuration
Speech_Synthesis_Models -- " Accesses Model Config " --> System_Utilities_Configuration
Velocity_Inference_Engine -- " Executes Model Inference " --> Speech_Synthesis_Models
Velocity_Inference_Engine -- " Returns Inference Output " --> Speech_Synthesis_Models
Velocity_Inference_Engine -- " Receives Engine Configuration " --> System_Utilities_Configuration
System_Utilities_Configuration -- " Provides Assets & Config " --> ChatTTS_Core_Orchestrator
System_Utilities_Configuration -- " Provides GPU & Config " --> Speech_Synthesis_Models
System_Utilities_Configuration -- " Provides Engine Config " --> Velocity_Inference_Engine
click ChatTTS_Core_Orchestrator href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main//ChatTTS/ChatTTS_Core_Orchestrator.md" "Details"
click Text_Processing_Module href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main//ChatTTS/Text_Processing_Module.md" "Details"
click Speech_Synthesis_Models href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main//ChatTTS/Speech_Synthesis_Models.md" "Details"
click Velocity_Inference_Engine href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main//ChatTTS/Velocity_Inference_Engine.md" "Details"
click System_Utilities_Configuration href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main//ChatTTS/System_Utilities_Configuration.md" "Details"
Browse more examples: GeneratedOnBoardings Repository
Codeboarding is integrated with everything we use:
- 📦 VS Code Extension: Interact with the diagram directly in your IDE.
- ⚙️ GitHub Action: Automate diagram generation in CI/CD.
- 🔗 MCP Server: Serves the consize documentation to your AI Agent assistant (ClaudeCode, VSCode, Cursor, etc.)
We’re just getting started and would love your help! If you have ideas, spot bugs, or want to improve anything - please open an issue or tackle an existing one. We actively track suggestions and welcome pull requests of all sizes.
Unified high-level representation for codebases that is accurate (hence static analysis). This representation is used by both people and agents → fully integrated in IDEs, MCP servers, and development workflows.