CodeRAG

CodeRAG is an enterprise-grade code intelligence platform that maps your codebases into a rich knowledge graph using Neo4j. It analyzes multiple languages to identify classes, methods, dependencies, architectural relationships, and project structure. It can scan remote repositories directly from GitHub, GitLab, or Bitbucket, including private repositories with secure authentication, and it provides semantic code search to answer natural-language queries like “show me where a specific API is used across projects.” The platform also surfaces quality metrics such as CK metrics and coupling patterns to help identify technical debt and opportunities for modernization. With multi-project management, CodeRAG enables enterprise teams to manage and compare several codebases from a single interface, making cross-project analysis and governance feasible. To get started, deploy the CodeRAG server (via the Docker image in this document), point it at your Neo4j instance, and connect your AI tooling or dashboards to leverage the semantic search, relationship mapping, and architectural analysis features.

Prerequisites:

• Docker installed on your host (Docker Engine and Docker CLI)
• Access to a Neo4j instance (URL, username, and password)

Installation steps:

1. Pull the CodeRAG Docker image (or build it if you have a custom image): docker pull coderag:latest

2. Run the CodeRAG container with required environment variables and a link to Neo4j: docker run -d --name coderag
   -e NEO4J_URL=bolt://neo4j:7687
   -e NEO4J_USERNAME=neo4j
   -e NEO4J_PASSWORD=password
   -p 8080:8080
   coderag:latest

3. Verify the service is running and accessible (adjust port mapping as needed). If you need to connect to a remote Neo4j instance, ensure network access and proper credentials are configured via the environment variables above.

4. Optional: configure remote repository access tokens and security settings as required by your environment. You can establish connectivity to GitHub/GitLab/Bitbucket using the platform’s built-in authentication mechanisms.

5. Connect your client tools or dashboards to the exposed API endpoints provided by CodeRAG to start scanning projects and querying the knowledge graph.

Tips and common issues:

• Ensure your Neo4j instance is reachable from the CodeRAG container. Network routing or VPNs may require additional configuration.
• For private repositories, provide a valid access token or SSH key configuration as required by your integration method.
• If you scale to multiple projects, use the multi-project management features to isolate analyses while enabling cross-project insights.
• The semantic search capability relies on embeddings; ensure your deployment has sufficient resources (CPU/Memory) to support indexing and query workloads.
• If you encounter connection errors to Neo4j, verify that the NEO4J_USERNAME and NEO4J_PASSWORD are correct and that the Bolt port (default 7687) is reachable from the container.
• Monitor container logs (docker logs coderag) for startup progress and error messages to diagnose issues quickly.