Installation
CodeGraphContext (CGC) is distributed as a Python package. You can install it using several methods depending on your workflow.
1. Install the CLI
Recommended: Using uvx (Fastest)
If you use uv, you can run CGC instantly without manual installation:
uvx codegraphcontext --help
Using pipx (Isolated)
For a persistent global installation in an isolated environment:
pipx install codegraphcontext
Using pip
pip install codegraphcontext
2. Database Backend Selection
CGC requires a graph database to store the indexed code. You can choose the backend that best fits your needs.
Option A: LadybugDB (Default & Recommended)
LadybugDB is an embedded, extremely fast graph database. It requires zero configuration and runs directly within the CGC process.
- Installation:
pip install real_ladybug - Best for: Local development, individual projects, and zero-ops setups.
- Pros: No external services, portable database files.
Option B: FalkorDB (High Performance)
FalkorDB is a low-latency graph database. CGC supports both local (embedded) and remote instances.
- Installation:
pip install falkordblite(Linux/macOS only) - Best for: Large codebases and performance-critical queries.
- Pros: Industry-leading query performance.
- Note: We use
falkordblitefor supported devices (Python 3.12+ on Unix), and LadybugDB (kuzudb) for the rest. We have largely shifted to LadybugDB as the primary embedded engine.
Option C: Neo4j (Enterprise)
Neo4j is the industry standard for graph databases, offering powerful visualization and management tools.
- Best for: Teams, massive repositories, and deep visual analysis via the Neo4j Browser.
- Setup: Requires a running Neo4j instance (Docker or Cloud).
codegraphcontext config set-db neo4j
3. Verify Installation
To ensure everything is configured correctly, run the version check:
codegraphcontext --version
You can also run the diagnostic command to check backend connectivity:
codegraphcontext doctor
4. Next Steps
Now that CGC is installed, you are ready to index your first repository.