Installation

CodeSight requires Python 3.10+ and an API key from at least one LLM provider (or Ollama for offline use).

From PyPI

Terminal

$ pip install codesight

From Source

Terminal

$ git clone https://github.com/AvixoSec/codesight.git
$ cd codesight
$ pip install -e ".[dev]"

Verify Installation

Terminal

$ codesight --version
codesight 0.3.0

Provider Setup

At least one provider must be configured. Fastest way:

Terminal

$ codesight config

Or set environment variables directly:

OpenAI

Terminal

$ export OPENAI_API_KEY="sk-..."

Anthropic

Terminal

$ export ANTHROPIC_API_KEY="sk-ant-..."

Google Vertex AI

Terminal

$ export GOOGLE_CLOUD_PROJECT="my-project"
$ gcloud auth application-default login

Ollama (fully offline)

Terminal

$ ollama serve
$ ollama pull llama3
$ codesight review my_file.py --provider ollama

No data leaves your machine with Ollama. Everything runs locally.

OpenAI-Compatible Providers (OpenRouter, Groq, Azure, etc.)

Any endpoint that speaks the OpenAI Chat Completions API - OpenRouter, Groq, Together, Mistral, xAI, Fireworks, DeepSeek, Perplexity, Cerebras, Cohere, Azure AI Foundry. Configure via the wizard:

Terminal

$ codesight config
  Select a provider: Custom (OpenRouter / Groq / Together / any OpenAI-compat)
  Pick a provider:   OpenRouter
  API key:           sk-or-v1-...
  Config label:      openrouter

$ codesight review file.py --provider openrouter

See the Providers page for the full preset list and config file format.

Check Connectivity

Terminal

$ codesight health
✓ Provider is reachable and healthy.