Skip to content
MCP server

VIVATLAS

Self-hosted catalogue of your skills, agents and MCP servers, searchable by meaning

Pricing
Free
Platforms
self-hosted, Web, Android
Runs
Remotely
Version
1.0.13
Listing updated
1 day ago

About VIVATLAS

VIVATLAS is a free MCP server for knowledge and memory. It runs as a hosted remote server, so there is nothing to install locally.

VIVATLAS is for people who collect tools and then cannot find them. Point it at GitHub or Gitea repositories and the links you clip, and it builds a card for each one with a name, a human-readable description, a picture and tags. A recommender can answer a task such as an accessibility audit with three candidates and a sentence on why each fits, or say nothing fits.

It runs on your own hardware as a Docker container, supports several users with private and shared catalogues, and only reads public repositories: it writes nothing back to Git and does not scan private repos. Descriptions and embeddings come from a Gemini key or a local Ollama server, and without AI the search falls back to words. Clippers exist for Chrome, Edge, Brave, Firefox and Android.

The MCP endpoint lets an assistant ask which of your tools does a given job. The server code is under the Business Source License 1.1.

Install

Run in a terminal

claude mcp add vivatlas --env SECRET_KEY='<SECRET_KEY>' --env GOOGLE_API_KEY='<GOOGLE_API_KEY>' -- docker run -i --rm -e SECRET_KEY -e GOOGLE_API_KEY ghcr.io/bobpanil/vivatlas:1.0.13

Add to claude_desktop_config.json

{
  "mcpServers": {
    "vivatlas": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "SECRET_KEY",
        "-e",
        "GOOGLE_API_KEY",
        "ghcr.io/bobpanil/vivatlas:1.0.13"
      ],
      "env": {
        "SECRET_KEY": "<SECRET_KEY>",
        "GOOGLE_API_KEY": "<GOOGLE_API_KEY>"
      }
    }
  }
}

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "vivatlas": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "SECRET_KEY",
        "-e",
        "GOOGLE_API_KEY",
        "ghcr.io/bobpanil/vivatlas:1.0.13"
      ],
      "env": {
        "SECRET_KEY": "<SECRET_KEY>",
        "GOOGLE_API_KEY": "<GOOGLE_API_KEY>"
      }
    }
  }
}

Add to .vscode/mcp.json

{
  "servers": {
    "vivatlas": {
      "type": "stdio",
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "SECRET_KEY",
        "-e",
        "GOOGLE_API_KEY",
        "ghcr.io/bobpanil/vivatlas:1.0.13"
      ],
      "env": {
        "SECRET_KEY": "<SECRET_KEY>",
        "GOOGLE_API_KEY": "<GOOGLE_API_KEY>"
      }
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "vivatlas": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "SECRET_KEY",
        "-e",
        "GOOGLE_API_KEY",
        "ghcr.io/bobpanil/vivatlas:1.0.13"
      ],
      "env": {
        "SECRET_KEY": "<SECRET_KEY>",
        "GOOGLE_API_KEY": "<GOOGLE_API_KEY>"
      }
    }
  }
}

Add to your client's MCP settings (most use this shape)

{
  "mcpServers": {
    "vivatlas": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "SECRET_KEY",
        "-e",
        "GOOGLE_API_KEY",
        "ghcr.io/bobpanil/vivatlas:1.0.13"
      ],
      "env": {
        "SECRET_KEY": "<SECRET_KEY>",
        "GOOGLE_API_KEY": "<GOOGLE_API_KEY>"
      }
    }
  }
}

Configuration

  • SECRET_KEYSigns sessions and encrypts stored tokens. Set once and keep it; left empty, the container generates one and keeps it in /data.
  • GOOGLE_API_KEYGemini key for the descriptions, search embeddings and picture choice. Optional; Ollama works too (OLLAMA_URL). Also settable in the admin panel.

What it does

Search by meaning
Combines full-text and vector search so you can describe a task instead of naming a tool.
Automatic cards
Generates a name, descriptions, tags and a cover image from each repository's README.
Recommender
Suggests three options with a reason each, or reports that nothing fits.
Clippers
Browser extension and Android share sheet for saving links.
Multi-user
Separate private cards and folders plus a shared catalogue.
Read-only on Git
Never writes to repositories or scans private ones.

Good for

  • Rediscover a scraper you starred months ago
  • Ask an assistant which of your catalogued tools audits accessibility
  • Keep a team catalogue of internal MCP servers and skills
  • Track which catalogued repos have gone stale upstream

Pricing

Free

Free to use and deploy, source under Business Source License 1.1.

VIVATLAS FAQ

How do I install it?

Run the ghcr.io/bobpanil/vivatlas Docker image on port 8710, then set up the owner at /setup.

Does it need an AI key?

Optional. A Gemini key or Ollama writes descriptions and embeddings; without it search goes by words.

Does it change my repositories?

No. It only reads public Git repositories and has no setting to write.

Is there a demo?

Yes, a public demo whose accounts are deleted after 24 hours and which runs without AI.

Is VIVATLAS free?

Yes. VIVATLAS is free to use.

How do I add VIVATLAS to Claude Code?

Run this command in your terminal: claude mcp add vivatlas --env SECRET_KEY='<SECRET_KEY>' --env GOOGLE_API_KEY='<GOOGLE_API_KEY>' -- docker run -i --rm -e SECRET_KEY -e GOOGLE_API_KEY ghcr.io/bobpanil/vivatlas:1.0.13

What are the best alternatives to VIVATLAS?

The closest alternatives to VIVATLAS are REA, Codebase Memory, GitHub, Serena MCP: the IDE for your agent and XERJ.

Alternatives to VIVATLAS

  1. REA

    MCP server

    CLI and MCP server that lets coding agents reverse engineer apps and binaries

    Runs locally
    • Open source
    • 464K GitHub stars
    • Developer tools
  2. Codebase Memory

    MCP server

    Local MCP server that builds a code knowledge graph for AI coding agents

    Runs locally
    • Open source
    • 48.7K GitHub stars
    • Knowledge and memory
  3. GitHub

    MCP server

    Official MCP server connecting AI assistants to GitHub repos, issues and workflows

    Not checked yet
    • Remote
    • Open source
    • 34.5K GitHub stars
    • Developer tools
  4. MCP toolkit giving coding agents symbol-level code retrieval and editing

    Runs locally
    • Open source
    • 31.5K GitHub stars
    • Developer tools
  5. XERJ

    MCP server

    Local search engine for AI agents with cited answers over your code and docs

    Runs locally
    • Open source
    • 19.2K GitHub stars
    • Knowledge and memory
  6. Ouroboros

    MCP server

    Spec-first layer that pins requirements and verifies results of AI coding agent work

    Runs locally
    • Open source
    • 6.4K GitHub stars
    • Developer tools