HelixDB

The HelixDB MCP server is a Rust-based implementation that powers AI-driven memory and relationship features on top of the HelixDB graph database. It exposes a suite of MCP tools for storing and querying memories (business, product, customer data), tracking customer interactions, and navigating directional data. You can perform keyword and semantic search, manage interactions, update memories with automatic index refresh, and discover relationships between customers and products or services. To run it, start the HelixDB instance first (default 127.0.0.1:6969), then launch the HelixDB MCP server binary. The server offers a consistent API with 22 tools, including create, update, query, and delete operations, plus specialized search and navigation capabilities. The CLI and client libraries can access tools like search_bm25, search_semantic, create_business_memory, create_customer_interaction, update_business_memory, and query_navigation, among others. Use the recommended embedding mode or provide your own embeddings depending on your HelixDB configuration, and rely on the built-in automatic search index maintenance when data changes.

Prerequisites:

• HelixDB server running (default: 127.0.0.1:6969)
• Rust 1.70+ installed
• Git to clone repository (if building from source)

Installation steps:

1. Clone the MCP server repository: git clone https://github.com/pilsertech/helixdb-mcp-server.git cd helixdb-mcp-server

2. Ensure HelixDB is reachable at the default endpoint (127.0.0.1:6969) or adjust configuration accordingly.

3. Build from source (Rust) cargo build --release

4. Run the server (release build) ./target/release/helix-mcp-server

5. (Optional) Use a pre-built Docker image docker pull ghcr.io/pilsertech/helixdb-mcp-server:latest docker run -i -p 8000:8000 ghcr.io/pilsertech/helixdb-mcp-server:latest

Prerequisite notes:

• If you use the embedding option, set up the HelixDB embedding provider and keys according to the README (OpenAI, Gemini, etc.).
• Verify network access to HelixDB at 127.0.0.1:6969 by curling the health endpoint: curl http://127.0.0.1:6969/health
• Ensure your Rust toolchain matches the required version (1.70+).

Tips and troubleshooting:

• If you see a “Failed to connect to HelixDB” error, confirm HelixDB is running and the endpoint/port in the mcpconfig.toml matches (127.0.0.1:6969 by default).
• Enable verbose debug logs by setting the environment variable RUST_LOG=debug before running the server.
• After changing MCP tool configurations, restart the MCP server so the LLM clients pick up the updated tools.
• The server supports automatic search index maintenance on updates. If search results seem stale after an update, verify that the update operation completed and try a short reindex if needed.
• For health checks, use curl http://127.0.0.1:6969/health and ensure the response indicates a healthy status.