Vector DB MCP
Give your coding agent fast local semantic and symbol search over your repo without standing up a separate vector stack.
Overview
Vector DB MCP is a MCP server for the Build phase that provides assembly-accelerated DuckDB code search, indexing, and symbol lookup for coding agents.
What is this MCP server?
- Assembly-accelerated DuckDB backend for code search and indexing
- Symbol lookup tuned for developer workflows
- Stdio MCP transport for Claude Code, Cursor, and compatible clients
- OCI image ghcr.io/kroq86/vector-db-mcp for repeatable installs
- Local-first search without a hosted vector SaaS dependency
- Server version 0.1.1
- Stdio transport via OCI identifier ghcr.io/kroq86/vector-db-mcp:0.1.1
- Three advertised capabilities: code search, indexing, symbol lookup
Community signal: 3 GitHub stars.
What problem does it solve?
Agents waste turns grepping huge repos or waiting on slow embeddings when you need reliable symbol and code search during active development.
Who is it for?
Indie developers shipping with Claude Code or Cursor who want a local, DuckDB-based code index MCP without managing a separate vector database service.
Skip if: Teams that need managed cloud vector search, multi-repo enterprise governance, or non-code document RAG as the primary use case.
What do I get? / Deliverables
After install, your agent can query a local DuckDB-backed index for faster code navigation and symbol resolution in the MCP tool loop.
- MCP-accessible code search and symbol lookup over indexed project code
- Local DuckDB-backed index usable from agent tool calls
- Repeatable stdio server registration in your agent config
Recommended MCP Servers
Journey fit
Code indexing and lookup happen while you wire agents and tooling into the product you are building. Fits the integrations shelf because it connects Claude Code or Cursor to an accelerated DuckDB-backed search layer in the dev loop.
How it compares
DuckDB code-index MCP integration, not a brainstorming or planning agent skill.
Common Questions / FAQ
Who is Vector DB MCP for?
Solo and indie builders using MCP-enabled agents who want faster in-repo code search, indexing, and symbol lookup on their machine or container.
When should I use Vector DB MCP?
Use it during Build when you are integrating agent tooling and need the model to look up symbols and search code through MCP instead of brittle full-file reads.
How do I add Vector DB MCP to my agent?
Register the stdio MCP server using the OCI package ghcr.io/kroq86/vector-db-mcp:0.1.1 in your Claude Code or Cursor MCP config, then expose its search and index tools to the agent.