Skip to content
MCP server

DEEPBOM

Local static audit of AI model files with hash-bound evidence for deployment reviews

Install
Pricing
Open source
Platforms
Web, CLI, API
Runs
Locally or Remotely
Version
2.5.0
Listing updated
1 day ago

About DEEPBOM

DEEPBOM is an open-source MCP server for ai models and llm tooling. It can run locally on your machine or as a hosted remote server.

Rather than running a model, DEEPBOM reads the serialized file and reports what can be established from it: graph structure, tensors, quantization, memory, compatibility and ML-BOM evidence. The public source covers TFLite, ONNX, GGUF, SafeTensors, Core ML and bounded ExecuTorch artifacts. Findings separate observed facts from predicted compatibility and from things a static audit cannot know, such as runtime behaviour.

The README is explicit that it is not a regulatory compliance verifier and does not replace clinical validation or risk management, and the site frames it around medical AI deployment. Model bytes stay on the device in the browser audit; optional telemetry is off unless enabled.

It is available as a web app, an npm CLI (npx deepbom audit), Node and Python SDKs, a local stdio MCP package at version 2.5.0, and a remote endpoint at deepbom.org/mcp. A repository-local Agent Skill can be installed for Codex or Claude Code.

Install

Run in a terminal

claude mcp add --transport http deepbom https://deepbom.org/mcp

Add to Settings → Connectors → Add custom connector

https://deepbom.org/mcp
Install in CursorOpens Cursor, which asks you to confirm. Or add it by hand:

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "deepbom": {
      "url": "https://deepbom.org/mcp"
    }
  }
}
Install in VS CodeOpens VS Code, which asks you to confirm. Or add it by hand:

Add to .vscode/mcp.json

{
  "servers": {
    "deepbom": {
      "type": "http",
      "url": "https://deepbom.org/mcp"
    }
  }
}

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

{
  "mcpServers": {
    "deepbom": {
      "serverUrl": "https://deepbom.org/mcp"
    }
  }
}

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

{
  "mcpServers": {
    "deepbom": {
      "type": "http",
      "url": "https://deepbom.org/mcp"
    }
  }
}

What it does

Local-only audit
Artifact bytes are analysed on the device and not uploaded.
Multi-format support
Reads TFLite, ONNX, GGUF, SafeTensors, Core ML and ExecuTorch files.
Evidence IR
A common representation family for artifact, model, weight and provenance evidence.
Hash-bound exports
Evidence reports are bound to the exact artifact by hash.
Change comparison
Compare model versions to review before and after differences.
Agent Skill
Install a skill for Codex or Claude Code with an integrate command.

Good for

  • Audit an ONNX or GGUF file before deployment
  • Review quantization and memory pressure of a model
  • Compare two model versions for a change-impact review
  • Reconcile an ML-BOM against the actual artifact
  • Give a coding agent a local analyzer for model files

Pricing

Open source

Prices change. Check deepbom.org before you buy.

DEEPBOM FAQ

Does my model get uploaded?

No. The browser audit and CLI work locally, and the README says model bytes are not uploaded.

Which formats are supported?

TFLite, ONNX, GGUF, SafeTensors, Core ML and bounded ExecuTorch.

Does it certify regulatory compliance?

No. The README says it is not a complete BOM validator or regulatory compliance verifier.

What does it need to run?

Node.js 20 or newer for the npx CLI.

Is DEEPBOM free?

Yes. DEEPBOM is open source, so you can use and self-host it for free.

How do I add DEEPBOM to Claude Code?

Run this command in your terminal: claude mcp add --transport http deepbom https://deepbom.org/mcp

What are the best alternatives to DEEPBOM?

The closest alternatives to DEEPBOM are REA, Claude Code Ultimate Guide, Agent Roast, Midplane and Leash.

Alternatives to DEEPBOM

  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. MCP server for searching the Claude Code Ultimate Guide from any MCP client

    Runs locally
    • Open source
    • 6.5K GitHub stars
    • AI models and LLM tooling
  3. Agent Roast

    MCP server

    Offline review that flags agent changes which weaken tests, assertions or CI

    Runs locally
    • Open source
    • 4 Hacker News points
    • Developer tools
  4. Midplane

    MCP server

    Postgres gateway for AI agents with table grants, column masking and approval for writes

    Runs locally
    • Open source
    • 5 GitHub stars
    • Databases
  5. Leash

    MCP server

    Review what your AI touched, approve risky actions and undo supported changes

    • 1 tool
    • Remote
    • Free
    • 1 GitHub star
    • Security