ModelsAtlas
Explore
Atlas / MCP servers / colemurray / Moondream

MoondreamSAFE

mcp/colemurray/moondream-1
MCP server by colemurray · mcp · mcp server · moondream · Audited 2026-10-08
⚑ Report / re-scanSource ↗

Moondream MCP Server in Python

Verdict
SAFE
Grade
B
Trust score
89 /100
Exposed tools
6 6r · 0w · 0d
Transport
stdio · streamable-http
License
MIT
Stars
47
01

Overview

From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.

A FastMCP server for Moondream, an AI vision language model. This server provides image analysis capabilities including captioning, visual question answering, object detection, and visual pointing through the Model Context Protocol (MCP).

Features

  • 🖼️ Image Captioning: Generate short, normal, or detailed captions for images
  • ❓ Visual Question Answering: Ask natural language questions about images
  • 🔍 Object Detection: Detect and locate specific objects with bounding boxes
  • 📍 Visual Pointing: Get precise coordinates of objects in images
  • 🔗 URL Support: Process images from both local files and remote URLs
  • ⚡ Batch Processing: Analyze multiple images efficiently
  • 🚀 Device Optimization: Automatic detection and optimization for CPU, CUDA, and MPS (Apple Silicon)

Installation

Prerequisites

  • Python 3.10 or higher
  • PyTorch 2.0+ (with appropriate device support)

Using uvx (Recommended for Claude Desktop)

# Run without installation
uvx moondream-mcp

# Or specify a specific version
uvx moondream-mcp==1.0.2

Install from PyPI

pip install moondream-mcp

Install from Source

git clone https://github.com/ColeMurray/moondream-mcp.git
cd moondream-mcp
pip install -e .

Development Installation

git clone https://github.com/ColeMurray/moondream-mcp.git
cd moondream-mcp
pip install -e ".[dev]"

Quick Start

Running the Server

# Using uvx (no installation needed)
uvx moondream-mcp

# Using pip-installed command
moondream-mcp

# Or run directly with Python
python -m moondream_mcp.server

Claude Desktop Integration

Add to your Claude Desktop configuration file:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json

Using uvx (Recommended)

{
"mcpServers": {
"moondream": {
"comm
Read from source at commit 809c5148331eOBSERVED · 2026-10-08
02

Connect

Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.

claude-code
claude mcp add moondream-mcp --env OPENAI_API_KEY=${OPENAI_API_KEY} -- uvx moondream-mcp
claude-desktop
{
  "mcpServers": {
    "moondream-mcp": {
      "command": "uvx",
      "args": [
        "moondream-mcp"
      ],
      "env": {
        "OPENAI_API_KEY": "${OPENAI_API_KEY}"
      }
    }
  }
}
03

Exposed tools (6)

6 read · 0 write · 0 destructive.

ToolRiskDescription
analyze_imageread
batch_analyze_imagesread
caption_imageread
detect_objectsread
point_objectsread
query_imageread
04

Trust audit

SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.

LayerWhat it checksResult
L0Provenance & inventoryPASS
L1Static analysis of the codePASS
L2Instruction surface (what it tells the agent)PASS
L3Class-specific surfacePASS
L4Behavioural (sandbox)SKIPPED

What the source does

Filesystem
none-observed
Network
declared (4 observation(s))
Shell
none-observed
Dependencies
not all pinned
Secrets in source
none-found

Findings (1)

LOWSupply chain · supply.unpinned · CWE-829, CWE-1357
examples/requirements.txt
agents, openai, python-dotenv, rich, click, requests, pathlib2
Why it matters. 7 requirement(s) not pinned with ==
Fix. pin exact versions

Gates applied: no_behavioural_pass.

Audited 2026-10-08 · audit v0.4.1 · source sha 809c5148331efull audit observations/trust-audit/mcp-server/colemurray__moondream-1.json · Report an issue / request a re-scan
05

Audit history

Every audit this server has had. A grade with a past is a grade somebody is still checking.

DateSourceVerdictGradeScoreChange
2026-10-08809c5148331eSAFEB89first audit
06

Questions

What is the Moondream MCP server?

Moondream MCP Server in Python

What tools does Moondream expose?

6 in total: 6 read-only, 0 that write, and 0 that can delete or overwrite. Every one is listed on this page with its risk.

Is Moondream safe to connect to an agent?

The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean server reads B.

What credentials does Moondream need?

It reads OPENAI_API_KEY from the environment. Give it a token scoped to the least it needs — an agent that can be talked into calling a tool can be talked into calling it with your credentials.

How does Moondream run?

It speaks stdio and streamable-http, so it runs as a local process your client starts. It is published on PyPI as moondream-mcp.

How current is this page?

The grade is for one exact copy of the source (809c5148331e), read on 2026-10-08. The repository is watched and re-audited when it changes.

Advertisement