Enzyme
Semantic search and pattern discovery across your Obsidian or markdown vault so agents cite the right notes while you research and plan.
Overview
Enzyme is a MCP server for the Idea phase that provides semantic search and pattern discovery across a local Obsidian or markdown vault via ENZYME_VAULT_PATH.
What is this MCP server?
- Semantic search and pattern discovery over Obsidian and markdown vaults
- ENZYME_VAULT_PATH points at your local vault directory
- Distributed as versioned .mcpb bundles (macOS arm64 and linux x86_64 listed)
- Stdio MCP under the Enzyme title—local-first, no cloud vault upload implied
- Knowledge MCP—not a generic web scraper or Notion OAuth integration
- Server version 0.3.3
- mcpb releases for macOS arm64 and linux x86_64 listed
- Config env ENZYME_VAULT_PATH
Community signal: 45 GitHub stars.
What problem does it solve?
Your agent cannot see the connections in a sprawling Obsidian vault when you only paste single files into chat.
Who is it for?
Indie builders who run a personal Obsidian or markdown knowledge base and want research-grade recall inside Claude Code or Cursor.
Skip if: Teams storing docs only in cloud SaaS with no local vault, or workflows needing live web search instead of private notes.
What do I get? / Deliverables
After you register Enzyme with your vault path, the agent retrieves semantically relevant notes and patterns without leaving the MCP session.
- Semantic retrieval over vault notes in agent workflows
- Pattern discovery across linked markdown content
- Local stdio MCP integration without mandatory cloud sync
Recommended MCP Servers
Journey fit
Personal knowledge bases feed earliest journey work—competitor notes, specs, and backlinks—before and alongside build, but research is where discovery queries dominate. Research matches semantic recall over vaults when you are synthesizing ideas, not deploying infra or running ad campaigns.
How it compares
Local vault semantic-search MCP—not a skills.sh markdown skill or public web RAG plugin.
Common Questions / FAQ
Who is Enzyme for?
It is for developers and founders who keep research and specs in Obsidian or markdown folders and want MCP-powered semantic search on that local vault.
When should I use Enzyme?
Use it during idea and research (and adjacent docs work) when you need the agent to surface related notes, themes, and patterns across many linked markdown files.
How do I add Enzyme to my agent?
Download the matching enzyme-mcp mcpb release for your platform, set ENZYME_VAULT_PATH to your vault root, and register the stdio server in your MCP client configuration.