@arvoretech/memory-mcp

MCP server for team memory — a persistent knowledge base with semantic search for AI-assisted development.

What is Team Memory?

Team memory captures accumulated knowledge over time: architectural decisions, coding conventions, incident learnings, domain knowledge, and gotchas. Unlike skills (which are prescriptive patterns), memories are descriptive facts the team has learned.

Tools

| Tool | Description | |------|-------------| | search_memories | Semantic search across all memories | | get_memory | Get full content of a specific memory | | add_memory | Create a new memory entry | | list_memories | List memories with optional filters | | archive_memory | Soft-delete a memory | | remove_memory | Permanently delete a memory |

Memory Categories

• decisions — Architectural Decision Records (ADRs)
• conventions — Team coding standards and preferences
• incidents — Past bugs, outages, and their root causes
• domain — Business domain knowledge and glossary
• gotchas — Known issues, quirks, and workarounds

Memory File Format

---
title: Use PostgreSQL for all services
category: decisions
date: 2024-06-01
author: joao.barros
tags: [database, architecture]
status: active
---

## Context
We needed to choose between PostgreSQL and MongoDB.

## Decision
PostgreSQL, because we need ACID transactions and complex joins.

## Consequences
- Migrations managed by Ecto and Prisma
- No dynamic schema flexibility

Semantic Search

Uses @xenova/transformers with the paraphrase-multilingual-MiniLM-L12-v2 model for local embeddings, stored in LanceDB (an embedded vector database in memories/.lancedb/). Supports Portuguese and English queries with cosine similarity search and metadata filtering.

Falls back to keyword search if the embedding model fails to load.

Environment Variables

| Variable | Default | Description | |----------|---------|-------------| | MEMORY_PATH | ./memories | Path to the memories directory | | MEMORY_EMBEDDING_MODEL | Xenova/paraphrase-multilingual-MiniLM-L12-v2 | HuggingFace model for embeddings |

Usage with hub.yaml

mcps:
  - name: team-memory
    package: "@arvoretech/memory-mcp"
    env:
      MEMORY_PATH: ./memories

Development

pnpm install
pnpm dev          # Run with tsx
pnpm build        # Compile TypeScript
pnpm test         # Run tests
pnpm test:cov     # Run tests with coverage