Memory PouchDB

This project adds new features to the Memory server offered by the MCP team. It allows for the creation and management of a knowledge graph that captures interactions via a language model (LLM). 🚀

This repo was forked fromhttps://github.com/BRO3886/mcp-memory-customwhich was a great starting point, thanks again for fixing timestamps. This repo's goal was more to fix the issue with an ever increasing json file for context.

To install Knowledge Graph Memory Server for Claude Desktop automatically viaSmithery:

npx -y @smithery/cli install @bneil/mcp-memory-pouchdb --client claude

git clone git@github.com:bneil/mcp-memory-pouchdb.git cd mcp-memory-pouchdb

{ "mcpServers": { "memory": { "command": "node", "args": ["/path/to/mcp-memory-pouchdb/dist/index.js"], "env": { "MEMORY_FILE_PATH": "/path/to/custom/memory.json", "POUCHDB_PATH": "/path/to/custom/pouchdb_directory", "DISABLE_MEMORY_FILE": "true" } } } }

The server will fail to start if either environment variable is not set. 🚫

- POUCHDB_OPTIONS: JSON string of additional PouchDB configuration options
- DISABLE_MEMORY_FILE: Set to "true" to disable saving to memory.json file (only use PouchDB for storage)

Add this to yourclaude_desktop_config.json/.cursor/mcp.jsonfile:

{ "mcpServers": { "memory": { "command": "node", "args": ["/path/to/mcp-memory-pouchdb/dist/index.js"], "env" : { "MEMORY_FILE_PATH":"/home/.../local_dbs/memory.json", "POUCHDB_PATH":"/home/.../local_dbs/pouchdb" } } } }

Follow these steps for each interaction: 0. Memory Initialization: - At startup, execute a read_graph function to initialize memory access - If memory is empty, create a default entity for the user with basic placeholder info - Run initialization sequence: read_graph → create_entities → read_graph 1. User Identification: - You should assume that you are interacting with default_user - If you have not identified default_user, proactively try to do so 2. Memory Retrieval: - Always begin your chat by saying only "Remembering..." and retrieve all relevant information from your knowledge graph - Always refer to your knowledge graph as your "memory" - Verify memory access is functioning properly 3. Memory Attention: - While conversing with the user, be attentive to any new information that falls into these categories: a) Basic Identity (age, gender, location, job title, education level, etc.) b) Behaviors (interests, habits, etc.) c) Preferences (communication style, preferred language, etc.) d) Goals (goals, targets, aspirations, etc.) e) Relationships (personal and professional relationships up to 3 degrees of separation) 4. Memory Update: - If any new information was gathered during the interaction, update your memory as follows: a) Create entities for recurring organizations, people, and significant events, add timestamps to wherever required. You can get current timestamp via get_current_time b) Connect them to the current entities using relations c) Store facts about them as observations, add timestamps to observations via get_current_time 5. Error Recovery: - If memory retrieval fails, execute read_graph function immediately - Log any memory access failures for debugging - Continue the conversation with best available information 6. Memory Validation: - Periodically verify that memory access is functional by checking for core entities - If validation fails, attempt reconnection via read_graph IMPORTANT: Provide a helpful and engaging response, asking relevant questions to encourage user engagement. Update the memory during the interaction, if required, based on the new information gathered (point 3).

To start the Knowledge Graph Memory Server, run:

The server will listen for requests via standard input/output.

The server exposes several tools that can be called with specific parameters:

- Get Current Time⏰
- Set Memory File Path📁
- Create Entities➕
- Create Relations🔗
- Add Observations📝
- Delete Entities❌
- Delete Observations🗑️
- Delete Relations🔗
- Read Graph📖
- Search Nodes🔍
- Open Nodes🔓

- Inspired by the Memory server from Anthropic
- Powered by PouchDB for robust data storage 💾

Engram is a hosted MCP server that provides reliable memory for AI agents:

Self-hosted MCP memory server for personal use and teams

Persistent memory, teams, and projects for AI agents. 76 MCP tools for storing, recalling, and sharing knowledge across sessions.

Provides searchable local storage for Claude conversation history, enabling context retrieval during sessions.

Deterministic O(1) memory for AI agents — local-first, MCP-native, with multi-agent speaker attribution and millisecond recall.

Integrates with Mem0.ai to provide persistent memory capabilities for LLMs, supporting cloud, Supabase, and local storage.

Enables Large Language Models to store and retrieve persistent memories with intelligent search capabilities.

A memory management system for AI assistants to store, retrieve, and manage user information using a local database.

Agent memory with git-like version control. Custom LLMs turn conversations into structured facts with automatic conflict detection - your agent sees how decisions evolved, not four contradictory text blobs. 80% token reduction vs RAG/graph systems. MCP + REST.

Persistent, inspectable memory for AI agents via hosted MCP and API. Supports recall, structured query, lineage, correction, and tenant-scoped remote memory.