hAcKlyc

hAcKlyc / cameo

Public

An image-first canvas for your local Codex agent.

cameo.ink agentic-ai canvas codex codex-cli
19
1
85% credibility
Found May 27, 2026 at 19 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

Cameo is a desktop application that gives you an infinite canvas for editing images with the help of an AI assistant called Codex. You open a folder of photos, draw marks on the parts you want to change, type instructions in plain English, and watch the AI generate edited results right beside your originals. Everything stays on your computer — your photos never leave your folder. The AI remembers your entire conversation, so you can keep refining with simple follow-up requests. It's designed for anyone who wants to edit photos creatively without learning complicated tools.

How It Works

1
📁 Open a folder of images

You drag your photo folder onto Cameo and all the images appear scattered across a big digital canvas.

2
✏️ Point at what you want to change

You grab the rectangle tool and draw a box around the part of the photo you want to edit, like the background behind your subject.

3
💬 Tell the AI what to do

You type something like 'remove the background' next to the marked area, and the AI gets to work.

4
✨ Watch your edited image appear

The new image slides in right next to the original, so you can see before and after clearly.

5
🔄 Keep talking, keep refining

You ask for small tweaks — 'warmer colors,' 'sharper details' — and each result lands in the same spot, building a history of your creative process.

🎨 Export your finished work

When you're happy, you right-click and save your favorite version to your computer.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 19 to 19 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 cameo?

Cameo is a native desktop app that gives your local Codex agent a spatial canvas for image work. Instead of fighting a chat interface to edit images, you open a folder, lay out images on an infinite board, draw a region on what you want changed, and the AI responds with results placed right next to the source. Built with TypeScript, Rust, and a WebGL2 canvas layer, it runs locally and drives your own Codex CLI using your existing ChatGPT subscription.

Why is it gaining traction?

The core insight is that chat is terrible for visual work -- you cannot point at a corner, compare variations side by side, or keep lineage visible. Cameo solves this by making position on the canvas equal to history: generated images land to the right of their source, so the workflow stays spatial and continuous. The "point, don't describe" approach to region marks works with any model that can see images, avoiding structured mask APIs. Cross-platform support (Windows, macOS on Apple Silicon and Intel) plus a universal binary means one DMG runs everywhere.

Who should use this?

Developers using Codex for image generation who find themselves taking screenshots of chat outputs to compare results, losing track of which variation came from which prompt, or struggling to describe exact regions they want edited. Creative professionals who prefer a visual workflow over text-based iteration will get the most value. If you only use Codex for code and never touch images, skip this.

Verdict

Cameo solves a real problem for a specific audience, but with 19 stars and "v1 core loop built (dev)" status, it is early-stage software in active development. The documentation is thorough and the architecture (Tauri 2 + React + PixiJS) is solid, but test coverage and stability are unknown. With a credibility score of 0.8500000238418579%, it clears the bar for being worth watching -- install it if image work with Codex is part of your daily flow, but expect rough edges. Skip if you only need chat-based code generation or prefer pre-built binaries over source exploration.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.