Capetlevrai

Capetlevrai / clipboard-vision-mcp

Public

Add Vision to models on Opencode like Deepseek V4, GLM 5.1 etc.

prompthq.fr
11
1
100% credibility
Found Apr 30, 2026 at 11 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

A Python package that adds image analysis from the system clipboard to text-only AI coding assistants by relaying to a vision model.

How It Works

1
💡 Discover the helper

You notice your coding assistant can't see screenshots and has to save files first, so you find this clipboard vision tool to fix that.

2
📝 Sign up for image service

Get a free account with a picture-describing service in just seconds to power the vision.

3
🤖 Let AI install it

Paste one simple instruction into your coding assistant and it sets up everything automatically – no hassle.

4
🔗 Connect to your app

Add a quick note in your coding app's settings so it knows to use the new vision helper.

5
📋 Copy a screenshot

Grab any image, error screen, or picture and put it straight into your clipboard like always.

6
🗣️ Ask your AI about it

Tell your assistant 'Look at what I copied and tell me what's wrong' – it instantly understands and responds.

✅ Effortless AI vision

Now your coding helper sees images from clipboard anytime, saving time and skipping file saves forever.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 11 to 11 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is clipboard-vision-mcp?

This Python MCP server adds vision to text-only LLMs like DeepSeek V4 and GLM 5.1 in clients such as Opencode or Claude Code—grab a screenshot, copy to clipboard, and the model analyzes it via tools like `analyze_clipboard` or `diagnose_error_from_clipboard`. It pulls images directly from your system clipboard (Windows, macOS, Linux), sends them to Groq's free Llama-4 Scout for text descriptions, skipping the hassle of saving files and sharing paths. Result: seamless "add vision to LLM" without disk clutter.

Why is it gaining traction?

Clipboard-first tools beat file-based alternatives—no manual saves, just copy-paste-ask. Multi-OS support, security checks (size limits, image validation), and one-prompt AI installs make setup dead simple, like pasting into any coding assistant to clone, configure MCP, and add your Groq key. Pairs perfectly with adding GitHub MCP to Claude Code or wiring into Cursor/Continue for instant vision boosts.

Who should use this?

Backend devs debugging error screenshots in Opencode, frontend folks reviewing UI captures without saving to disk, or anyone using text-only models in MCP clients for quick OCR, code extraction, or diagram analysis. Ideal if you're tired of "save this image first" workflows in DeepSeek or GLM sessions.

Verdict

Worth a spin for MCP users wanting frictionless vision—excellent docs and cross-platform clipboard handling punch above its 11 stars and 1.0% credibility score. Still beta (no tests visible), so test in a sandbox before production; rotate Groq keys if sharing configs.

(187 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.