Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@0f recall why we chose gRPC for inter-service communication"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
The universal, local-first memory hub and Model Context Protocol (MCP) daemon for AI coding agents.
English · Português · Español · Français · Deutsch · Русский · 中文 · 日本語 · 한국어
Web Dashboard · MCP Setup Guide · Architecture · Charter
When you switch between AI coding assistants, you shouldn't have to re-explain your architecture, re-teach your rules, or lose track of past decisions.
Zero Friction (0f) provides a single, local-first daemon and Model Context Protocol (MCP) server so all your AI tools share the same brain:
┌──────────────────────────────────────────────┐
│ Your Developer Machine │
└──────────────────────┬───────────────────────┘
│
┌───────────────────┬─────────────┴───────┬───────────────────┐
▼ ▼ ▼ ▼
[ Claude Code ] [ Cursor IDE ] [ Antigravity / AGY ] [ Windsurf ]
│ │ │ │
└───────────────────┴─────────────┬───────┴───────────────────┘
│ Standard MCP (stdio)
▼
┌──────────────────────────────────────────────┐
│ Zero Friction (0f) │
├──────────────────────────────┬───────────────┤
│ 🧠 LanceDB Hybrid Memory │ 🕸️ SQLite │
│ (Dense Vectors + BM25) │ Knowledge │
│ │ Graph │
└──────────────────────────────┴───────────────┘Related MCP server: Agent Memory Bridge
┌─────────────────────────────────────────────────────────────┐
│ Zero Friction Storage Engine │
├─────────────────┬─────────────────────────┬─────────────────┤
│ Tier 1: Local │ Tier 2: Teams (Git) │ Tier 3: AWS │
│ (Individual) │ (Engineering Teams) │ (Enterprise) │
├─────────────────┼─────────────────────────┼─────────────────┤
│ • Zero setup │ • Git repo sync │ • AWS S3 Arrow │
│ • Local LanceDB │ • Automatic Markdown ADR│ • DynamoDB Graph│
│ • ~/.0f/ storage│ • Audit trail via PRs │ • KMS Encryption│
└─────────────────┴─────────────────────────┴─────────────────┘Tier 1 (Individual / Local-First): Stored locally in ~/.0f/ with LanceDB & SQLite. 100% private, zero setup.
Tier 2 (Teams / Git-Backed): Sincronize memories across your team using any private Git repository (0f sync --tier git --remote <git_url>). Generates human-readable Markdown ADRs automatically.
Tier 3 (Enterprise / Managed Cloud): High-scale AWS cloud storage plugin (0f cloud-init aws) using S3 for vector partitions with KMS encryption and DynamoDB/Aurora for graph topologies.
0f Command Line Tool# 1. Store an architectural decision (ADR) or memory
0f remember "ADR-001: Adopt gRPC" "We choose gRPC for low-latency microservice RPC." --tag grpc --tag architecture
# 2. Recall context semantically across your memories
0f recall "how do we communicate between microservices?"
# 3. Inspect storage tiers (Tier 1 Local, Tier 2 Git, Tier 3 AWS)
0f tiers
# 4. Sync memories with your team's Git repo
0f sync --tier git --remote git@github.com:myorg/team-memory.git
# 5. Export AWS Terraform self-provisioning module
0f cloud-init aws --dest ./terraform-0f
# 6. Inspect knowledge graph topology
0f graph
# 7. Check local hub health & storage
0f status
# 8. Start Model Context Protocol (MCP) server
0f mcpAdd 0f to your claude_desktop_config.json or Cursor MCP settings:
{
"mcpServers": {
"0f": {
"command": "0f",
"args": ["mcp"]
}
}
}Now any AI agent can call remember_context, recall_context, and list_adrs automatically!
Distributed under the Apache 2.0 License. See LICENSE for details.
This server cannot be deployed
Hosted MCP memory for coding agents: persistent across sessions, editable markdown, team sharing.
shared AI-context layer for teams — persistent memory your agents search and update over MCP
Persistent AI memory shared across Claude, ChatGPT, coding agents, and compatible MCP clients.
Shared memory for coding agents. Stop re-explaining your codebase every session.
