mcpField
V
vibo-mcp

vibo-mcp

Local
github.com/vnbochkarev-netizenMemory & Knowledge1Updated 6w ago
Repository

About vibo-mcp

vibo-mcp is a persistent-memory MCP server for AI agents, with encrypted memory levels, thread memory, and tools to search and save facts. It also reports token-savings statistics and is designed to work with MCP clients.

Key features

  1. 01Search and save persistent facts
  2. 02Deduplicate saved facts by exact match
  3. 03Organize memory across L1/L2/L3
  4. 04Add, compress, query, and contextualize threads
  5. 05View token-savings statistics

Use cases

  1. 01Recall stored preferences across agent sessions
  2. 02Save facts for later retrieval
  3. 03Compress and query ongoing thread context

MCP support

Provides MCP tools for memory search, saving, usage statistics, and thread memory. It is configured in MCP clients using a VIBO_API_KEY.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

mcp.json · Universal config
{
  "mcpServers": {
    "vnbochkarev-netizen-vibo": {
      "command": "npx",
      "args": [
        "-y",
        "@vibo-dev/vibo-mcp"
      ],
      "env": {
        "VIBO_API_KEY": "<VIBO_API_KEY>"
      }
    }
  }
}
Run commandnpx -y @vibo-dev/vibo-mcp
Environment variables to set
VIBO_API_KEYSecretYour ViBo license key (VIBO-...). Get a free trial: https://wwwvibo.com

Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.

Packages
npm@vibo-dev/vibo-mcp0.2.6

Tools

6

Saved tool list · updated 2w ago

ParameterTypeRequiredDescription
limitnumberNo
querystringYesWhat to search for

Details

From the project README

🦞 ViBo MCP Server

Memory for AI agents — persistent memory (L1/L2/L3 encryption), web-search savings and thread memory. Works with any MCP client: Claude Desktop, Cursor, OpenClaw, Windsurf, Codex, and more.

Install

npm install -g @vibo-dev/vibo-mcp

Configuration

Add to your MCP client config (Claude Desktop example):

{
  "mcpServers": {
    "vibo": {
      "command": "npx",
      "args": ["-y", "@vibo-dev/vibo-mcp"],
      "env": {
        "VIBO_API_KEY": "YOUR_VIBO_KEY"
      }
    }
  }
}

Get a key: https://wwwvibo.com — free 2-day trial, then $5/month.

Tools

ToolDescription
memory_searchFind relevant facts (returns token savings)
memory_addSave a fact (dedup by exact match)
memory_usageYour real savings statistics
thread_memoryThread: add / compress / ask / context

Example

Agent: "What does client Anna prefer?" → memory_search("Anna preferences") → • [L1] client-anna: Anna likes coffee without sugar, order #42 → 💾 Saved 13,452 tokens (97.5%)

Honest numbers

  • Memory: 97.5% fewer tokens on 118 facts (grows with memory, 50-150× on 10K+).
  • Web search: 99.6% (measured 47,443 → 186 tokens).
  • Threads: -72%.
  • Secrets (L3) never reach the LLM — encrypted by design.

Links

Related projects

  • CloudArc — pack a 10 GiB folder into one .vibo archive at a flat ~26 MiB peak RSS, and read its index over HTTP without fetching the payload (Apache-2.0). Docs site
  • memory-shield — poisoning defense for agent memory (MIT).
  • ViBo-memory — persistent agent memory with L1/L2/L3 encryption.

Integrations

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.