Automating Repository Documentation Privately with CodeWiki

Python, Java, and JavaScript
TypeScript, C, and C++
C#, Kotlin, and PHP

Getting started is straightforward via pip. You can install the tool directly from the source to ensure you have the latest research-backed features. For those managing complex environments, you can follow the official CodeWiki installation guide.

# Install via pip
pip install git+https://github.com/F0Soft-AI4Code/CodeWiki.git

# Generate docs for the current project with a browsable HTML viewer
codewiki generate --github-pages --create-branch

# Only regenerate what changed since the last run
codewiki generate --update

Maintaining up-to-date architectural documentation is a constant battle for engineering teams. While cloud-based tools like Google Code Wiki offer convenience, they often require you to ship your proprietary source code to external servers.

Enter CodeWiki, an open-source Python CLI designed for holistic, architecture-aware documentation. Unlike many hosted solutions, it runs entirely on your local machine, ensuring that your private repositories never leave your control.

Privacy-First Documentation Engine

CodeWiki is more than just a simple README generator. It analyzes your codebase to produce high-level Mermaid diagrams, including architecture, data-flow, and sequence diagrams, alongside detailed module-level prose.

Feature	Hosted AI Tools	CodeWiki (Local)
Data Privacy	Source code sent to cloud	Runs on your local machine
Infrastructure	SaaS / Remote execution	Local CLI execution
Cost Model	Per-token usage fees	Local LLM or subscription mode

The tool supports a wide range of environments, making it versatile for polyglot engineering teams. It provides deep analysis across nine major programming languages:

Python, Java, and JavaScript
TypeScript, C, and C++
C#, Kotlin, and PHP

Implementation and Usage

# Install via pip
pip install git+https://github.com/F0Soft-AI4Code/CodeWiki.git

# Generate docs for the current project with a browsable HTML viewer
codewiki generate --github-pages --create-branch

# Only regenerate what changed since the last run
codewiki generate --update

One of the most powerful features is its subscription mode. This allows you to route LLM calls through your local Claude or Codex CLI, effectively letting you use your existing AI plans rather than paying additional per-token fees for documentation generation.

Evaluating the Research Artifact

It is vital to approach CodeWiki with a critical eye. As a research artifact (sponsored by FPT Software and slated for an ACL 2026 paper), it is an experiment you should validate rather than a hardened production tool. While the authors report that it outperforms DeepWiki in their internal benchmarks, their own data shows a performance dip when handling systems languages like C and C++.

Furthermore, because it performs full LLM calls to generate high-quality diagrams, you must account for the token costs of your chosen model. For those looking to scale this locally, you can explore the CodeWiki MCP implementation or use the Docker deployment documentation for containerized setups.

For a robust, network-free deployment strategy, refer to this configuration validation workflow. You can also find the complete project details at the official CodeWiki repository.

Verdict

If you are working on sensitive, private codebases and need architecture-level visibility without the privacy risks of cloud-native tools, CodeWiki is a highly promising experiment worth testing in your local workflow.

Ready to secure your documentation pipeline? Try running CodeWiki on your next local Python project today!

CLI Commands

Configuration Management

# Set up your API configuration
codewiki config set \
  --api-key <your-api-key> \
  --base-url <provider-url> \
  --main-model <model-name> \
  --cluster-model <model-name> \
  --fallback-model <model-name>

# Configure max token settings
codewiki config set --max-tokens 32768 --max-token-per-module 36369 --max-token-per-leaf-module 16000

# Configure max depth for hierarchical decomposition
codewiki config set --max-depth 3

# Show current configuration
codewiki config show

# Validate your configuration
codewiki config validate

Documentation Generation

# Basic generation
codewiki generate

# Custom output directory
codewiki generate --output ./documentation

# Create git branch for documentation
codewiki generate --create-branch

# Generate HTML viewer for GitHub Pages
codewiki generate --github-pages

# Enable verbose logging
codewiki generate --verbose

# Full-featured generation
codewiki generate --create-branch --github-pages --verbose

# Incremental update (only regenerate changed modules since last run)
codewiki generate --update

# Incremental update using a specific commit hash to compare against (useful in CI/CD or squashed PRs)
# This overrides the stored commit hash in metadata.json and implicitly enables --update
codewiki generate --compare-to <commit-hash>

Privacy-First Documentation Engine

Feature	Hosted AI Tools	CodeWiki (Local)
Data Privacy	Source code sent to cloud	Runs on your local machine
Infrastructure	SaaS / Remote execution	Local CLI execution
Cost Model	Per-token usage fees	Local LLM or subscription mode

The tool supports a wide range of environments, making it versatile for polyglot engineering teams. It provides deep analysis across nine major programming languages: