Skip to main content
MCP server CX Codex

linksee-memory — Claude Code MCP server: setup and guide

Local SQLite memory MCP server that passes architecture decisions between agent sessions and flags drift in your codebase.

Install into Claude Code · Platform CLI All methods and platforms ↓
claude mcp add linksee-memory -- npx -y linksee-memory
Open on GitHub

What linksee-memory does

The server logs project decisions and caveats inside a local SQLite database. When starting a fresh session, the agent retrieves active guidelines using recall(). The tool surfaces locked constraints before file edits or write actions occur, preventing the model from reintroducing discarded patterns.

Built-in drift tracking matches code against recorded decisions using visual status indicators. Several environments like Claude Code and Cursor share the identical database. Everything runs strictly on your machine. No external cloud credentials are required.

What it helps with

  • retain architecture decisions across fresh Claude Code sessions
  • prevent agents from re-implementing rejected ideas
  • share persistent decisions across Claude Code and Cursor
  • audit codebase drift against written specifications

Example prompts for the agent

  • Check linksee and show active caveats before modifying the database schema
  • Remember this architectural constraint in linksee so future sessions do not revert it
  • Check what is drifting between current code and locked decisions

How to install linksee-memory

Platform CLI

claude mcp add linksee-memory -- npx -y linksee-memory

Step by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more

Questions about linksee-memory

How to install linksee-memory in Claude Code?

Run claude mcp add linksee-memory -- npx -y linksee-memory in your terminal.

What does linksee-memory do?

It preserves engineering decisions in a local SQLite file and injects them back into agent sessions before files are modified.

Does linksee-memory require remote API keys?

No. The server stores all data locally in an SQLite database without calling third-party cloud services.

Similar MCP servers

All MCP servers for Agent setup: memory, context, orchestration →

Project sponsors

Companies that keep this analytics open
Gold sponsor

Impressions, clicks and code copies — open statistics. We keep the numbers public for the same reason NeBlask itself is public. Become a sponsor