AI & Machine Learning

deepbom.org

Analyzes and validates machine learning deployment artifacts such as TFLite, ONNX, GGUF, SafeTensors, Core ML, and ExecuTorch files, and manages DEEPBOM analysis results and errors.

2 endpoints8 known toolsFirst detected September 9, 2026Last detected September 27, 2026

ENDPOINT 1

https://deepbom.org/mcp

No auth detected

MCP server metadata

Name: deepbom-chatgpt
Version: 1.108.0

Capabilities

toolsresources

Server instructions

Use deepbom_analyze_file when a user attaches a supported serialized AI deployment artifact and asks for static inspection, a cross-format model summary, or a model-structure visualization. The attachment is analyzed in the ChatGPT browser sandbox; the DEEPBOM server receives only a bounded result, not model bytes. Do not present a result until deepbom_publish_analysis returns it. Report model_summary rows as serialized operations or storage, not framework trainability or runtime order. The widget renders deterministic Model IR views locally and can export SVG, PNG, and a Word-ready bundle; do not reconstruct a graph from the bounded conversation summary or invent details absent from it. Keep artifact_defect, caution, and evidence_gap distinct. Static evidence does not establish measured latency, accuracy, clinical validity, regulatory compliance, standard conformance, or actual runtime placement. Use the local DEEPBOM MCP for confidential files or artifacts too large for ChatGPT.

Known tools 4

deepbom_analyze_file

Use when the user asks to inspect, summarize, or visualize an attached TFLite, ONNX, GGUF, SafeTensors, Core ML, or ExecuTorch deployment artifact.

Inferred read-only

deepbom_capabilities

Use before analysis when format support, privacy boundaries, or the difference between the ChatGPT and local DEEPBOM paths is unclear.

Inferred read-only

deepbom_publish_analysis

Receives one bounded browser-produced DEEPBOM result, validates its structure, and returns it to the conversation.

Inferred read-only

deepbom_publish_error

Receives one bounded browser-produced DEEPBOM error and returns a recovery action to the conversation.

Inferred read-only

CONNECT WITH APPROVAL

Client installation

Review this server and its permissions before adding it. Secret placeholders must be set locally.

Codex

~/.codex/config.toml

[mcp_servers.deepbom-chatgpt]
url = "https://deepbom.org/mcp"
enabled = true

Claude Code

.mcp.json

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

Claude Desktop

Settings → Connectors → Add custom connector

Name: deepbom-chatgpt
Remote MCP URL: https://deepbom.org/mcp

Add this remote URL as a custom connector in Claude Desktop. Availability depends on the user plan and workspace policy.

Cursor

.cursor/mcp.json

{
  "mcpServers": {
    "deepbom-chatgpt": {
      "url": "https://deepbom.org/mcp"
    }
  }
}

Visual Studio Code

.vscode/mcp.json

Add to Visual Studio Code

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

Generic MCP

Client-specific MCP configuration

{
  "name": "deepbom-chatgpt",
  "transport": "streamable-http",
  "url": "https://deepbom.org/mcp"
}

MCP Inspector

Run the official MCP Inspector locally and enter the indexed Streamable HTTP endpoint.

ENDPOINT 2

https://deepbom.org/mcp/claude

No auth detected

MCP server metadata

Name: deepbom-claude-browser
Version: 1.108.0

Capabilities

toolsresources

Server instructions

This connector opens an MCP App where the user deliberately selects one supported deployment artifact. It cannot automatically read Claude attachments or arbitrary local files. Analysis runs in the browser sandbox; the service receives a bounded evidence result, not model bytes. The model_summary field is a Common Model IR table: preserve its trainability and ordering boundaries and do not reinterpret storage as framework parameters. For confidential, large, sharded, or package artifacts, use the local DEEPBOM desktop extension or CLI.

Known tools 4

deepbom_open_local_analyzer

Opens an interactive browser-local analyzer where the user selects one supported serialized AI deployment artifact.

Inferred read-only

deepbom_capabilities

Returns supported formats, transfer boundaries, and the distinction between the remote browser-local workflow and the local desktop extension.

Potential side effects

deepbom_publish_browser_analysis

Validates and returns one bounded result created in the MCP App sandbox.

Inferred read-only

deepbom_publish_browser_error

Validates and returns a bounded browser-local diagnostic with a recovery action.

Inferred read-only

CONNECT WITH APPROVAL

Client installation

Review this server and its permissions before adding it. Secret placeholders must be set locally.

Codex

~/.codex/config.toml

[mcp_servers.deepbom-claude-browser]
url = "https://deepbom.org/mcp/claude"
enabled = true

Claude Code

.mcp.json

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

Claude Desktop

Settings → Connectors → Add custom connector

Name: deepbom-claude-browser
Remote MCP URL: https://deepbom.org/mcp/claude

Add this remote URL as a custom connector in Claude Desktop. Availability depends on the user plan and workspace policy.

Cursor

.cursor/mcp.json

{
  "mcpServers": {
    "deepbom-claude-browser": {
      "url": "https://deepbom.org/mcp/claude"
    }
  }
}

Visual Studio Code

.vscode/mcp.json

Add to Visual Studio Code

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

Generic MCP

Client-specific MCP configuration

{
  "name": "deepbom-claude-browser",
  "transport": "streamable-http",
  "url": "https://deepbom.org/mcp/claude"
}

MCP Inspector

Run the official MCP Inspector locally and enter the indexed Streamable HTTP endpoint.

TRUST AND VERIFICATION EVIDENCE

Loading Trust v2 evidence…

Checking the associated registrable domain. The BuiltWith key remains server-side.

Indexed

Evidence is source-attributed and does not guarantee that a third-party server is safe. Risk labels are conservative metadata heuristics.