Skip to content

prerequisites.md

Latest commit

 

History

History
76 lines (53 loc) · 4.14 KB

prerequisites.md

File metadata and controls

76 lines (53 loc) · 4.14 KB
Raw

Prerequisites

Before setting up CodeWiki, make sure your environment meets the following requirements.

Required Software

Software Purpose Notes
Python 3 Runs the CLI and the FastAPI web application CodeWiki is a pure Python project (codewiki/ package)
pip Installs Python dependencies Standard with most Python installations
Git Repository cloning, branch/commit automation (GitManager), and repository validation Required for both the CLI's --create-branch workflow and the web app's repository cloning
Docker & Docker Compose (optional) Runs the web application as a container Only needed if you use docker/docker-compose.yml instead of running Python directly

CodeWiki analyzes source files across several languages (Python, JavaScript, TypeScript, Java, C, C++, C#, PHP) but CodeWiki itself is implemented entirely in Python — you do not need compilers or runtimes for those languages to run CodeWiki.

System Requirements

  • A machine capable of running Python 3 and cloning git repositories.
  • Outbound network access to your configured LLM provider endpoint(s), since documentation generation makes HTTP calls to LLM APIs (OpenAI-compatible endpoints, Anthropic, or a local LiteLLM proxy).
  • Sufficient disk space to clone target repositories and store generated documentation under an output/ directory.

Account / Access Requirements

  • LLM provider credentials — CodeWiki requires API keys for its LLM-backed generation stages. The pipeline is split into three provider "slots" — cluster, main, and fallback — each with its own model, API key, and base URL. You need at least one working provider configured for all three slots (they can point to the same provider/model).
  • GitHub access (optional) — If you plan to use codewiki generate --create-branch and open a pull request with generated docs, you'll need push access to the target repository.

Environment Variables

CodeWiki reads configuration either from CLI-managed local storage (~/.codewiki/config.json + system keyring) or from environment variables (used by Config.from_args() for the web app / non-interactive contexts). The core variables are:

Variable Required Description
MAIN_MODEL Optional (has default) Model used for main documentation generation. Defaults to claude-sonnet-4.
CLUSTER_MODEL Optional Model used for the module-clustering stage. Defaults to MAIN_MODEL.
FALLBACK_MODEL Required (via from_args) Model used when the main model fails.
CLUSTER_API_KEY Required (via from_args) API key for the clustering-stage provider.
MAIN_API_KEY Required (via from_args) API key for the main generation-stage provider.
FALLBACK_API_KEY Required (via from_args) API key for the fallback-stage provider.
LLM_BASE_URL Optional Base URL for the LLM endpoint. Defaults to http://0.0.0.0:4000/ (e.g. a local LiteLLM proxy).
MAX_OUTPUT_TOKENS Optional Maximum output tokens per LLM call. Defaults to 16384.
APP_PORT Optional Port the web application listens on when run via Docker Compose. Defaults to 8000.

Note: When using the CLI, most of these values are instead collected interactively and persisted via codewiki config set, with API keys stored securely in your OS keyring rather than plain environment variables. See codewiki config set --help for the full list of per-provider options (cluster/main/fallback model, base URL, max tokens, temperature, API version).

Verification Commands

Run the following to confirm your environment is ready:

# Confirm Python is installed
python3 --version

# Confirm pip is available
pip --version

# Confirm Git is installed
git --version

# (Optional) Confirm Docker and Docker Compose are installed
docker --version
docker compose version

Once CodeWiki is installed (see Quick Start), verify the CLI is working:

# Show CLI help
codewiki --help

# Show version info
codewiki version

Next Steps

Continue to the Quick Start guide to install CodeWiki and generate your first documentation set.