remindb

Overview

Remindb functions as an MCP server that transforms unstructured notes into a structured, indexed memory system. It parses various formats (Markdown, HTML, JSON, YAML, TOON) into a tree of typed nodes, each with temperature tracking, token optimization, and versioning capabilities. The system exposes a suite of MCP tools including MemoryTree for browsing, MemorySearch for fast FTS5 search, MemoryDelta for incremental updates, and MemoryRelated for knowledge graph navigation.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

• MemoryTree

  Get a tree view of the memory structure with typed nodes and temperatures

  Try:Show me the memory tree structureTry:What's the hot content in my notes right now?
• MemorySearch

  Search the memory using FTS5 with ranked results and token budgeting

  Try:Find all information about payment processingTry:Search for documentation on the data model
• MemorySummarize

  Summarize cold nodes to reduce token usage while preserving structure

  Try:Summarize the cold parts of my project documentationTry:Compact this stale information to save context
• MemoryDelta

  Get changes since last cursor for efficient resync

  Try:What's new in my notes since last time?Try:Show me the updates to the documentation
• MemoryRelated

  Find knowledge graph relations between nodes up to 5 hops

  Try:What's related to the architecture section?Try:Show me content connected to payment integration
• MemoryRelate

  Create new knowledge graph relationships between nodes

  Try:Connect the data model section to the schema documentationTry:Link this observation to the relevant user feedback

Comparable tools

Installation

curl -fsSL https://raw.githubusercontent.com/radimsem/remindb/main/install.sh | bash

iwr -useb https://raw.githubusercontent.com/radimsem/remindb/main/install.ps1 | iex

git clone https://github.com/radimsem/remindb.git
cd remindb
go build -o ~/.local/bin/remindb ./cmd/remindb

{
  "mcpServers": {
    "remindb": {
      "command": "remindb",
      "args": ["serve"]
    }
  }
}

FAQ

How does remindb reduce token usage?

By parsing notes into a structured tree with hash-based deduplication and temperature-based relevance, allowing agents to only access relevant portions of memory rather than re-reading entire files.

What file formats does remindb support?

It supports Markdown, HTML, JSON, YAML, and the TOON format, with automatic compression where beneficial.

Compare remindb with