ReviseMy
Open-source human-in-the-loop design review tool that enables AI agents to upload UI screenshots for human feedback, with MCP tools for creating reviews, pinning notes, and getting structured feedback.
README
ReviseMy
Mark feedback for your agent.
ReviseMy is an open-source human-in-the-loop design review tool. Your agent uploads UI screenshots over Laravel MCP, you open a review link, mark what matters like a design critique, then approve or request changes. The agent reads structured work packets and keeps going.
Built with Laravel, Livewire, Flux, Sanctum, and Laravel MCP — ready for Laravel Cloud.
Features
- Marks, not pins — product UI speaks in marks (rose rectangles + M1/M2 badges). Human marks are authoritative; API keys stay
pinsfor compatibility. - Rectangle-first review — drag to outline a region or click for a point note; zoom with +/− and pan with Space+drag (or middle mouse).
- Second opinion (hints only) — Cloud-queued design checklist on every screenshot; optional OpenAI vision when keyed. Sky S-markers never override your marks.
- Agent subagent path —
add_findingsdrops suggestion / a11y / polish notes into the same review before you look. - Work packets +
next_action— agents know whether to wait, apply marks, open the next pass, or stop. - Multi-pass checkups —
create_reviewwithparent_idfor pass 2+ after you request changes. - Try token, no account — one-click token on the homepage; paste MCP config for Cursor, Claude Code, VS Code, or ChatGPT.
- Secret review links — humans open
/r/{token}without signing up.
Try it on any project (~2 minutes)
- Open the hosted app (your
*.laravel.cloudURL after deploy). - Click Get a try token.
- Copy the MCP config for your client (Cursor, Claude, VS Code, or ChatGPT).
- Ask your agent to screenshot UI work and call
create_review. - Open the review link, mark feedback, approve or request changes.
- Ask the agent to call
get_reviewand follownext_action.
No account required for the human reviewer.
MCP tools
| Tool | Purpose |
|---|---|
create_review |
title + screenshots (+ optional page_url, parent_id for next pass) → review URL; queues second opinion |
get_review |
work packets + next_action (wait_for_human / apply_pins_then_next_pass / done) |
list_reviews |
recent reviews for this try token |
add_screenshot |
append a shot to an open review |
add_findings |
agent subagent — push suggestion/a11y/polish into the review |
request_second_opinion |
re-queue Cloud checklist (+ vision if keyed) |
Prompt: design_checkup_loop — full agent↔human checkup cycle.
Screenshots accept https URLs, data URLs, or base64.
Terminology: UI copy uses marks; JSON still uses work_packets.pins, related_pin, and apply_pins_then_next_pass. Second opinion is suggestions only — see docs/SECOND-OPINION.md.
MCP config (Cursor example)
After you get a try token from the homepage:
{
"mcpServers": {
"revisemy": {
"url": "https://YOUR-APP.laravel.cloud/mcp/revisemy",
"headers": {
"Authorization": "Bearer YOUR_TOKEN"
}
}
}
}
The homepage also copies configs for Claude Code, VS Code (servers key), and ChatGPT. Same URL + Bearer header works for any MCP host — see docs/CONNECTORS.md.
REST API (same auth)
POST /api/try-token— create a try workspace + tokenPOST /api/reviews—{ "title", "context?", "page_url?", "parent_id?", "images": [...] }GET /api/reviews/{id}GET /api/reviewsPOST /api/reviews/{id}/screenshots—{ "image" }POST /api/reviews/{id}/findings—{ "findings": [...] }POST /api/reviews/{id}/second-opinion— optional{ "screenshot_index" }
Local development
composer install
cp .env.example .env
php artisan key:generate
touch database/database.sqlite
php artisan migrate
php artisan storage:link
npm install && npm run build
php artisan serve
Visit http://127.0.0.1:8000, get a try token, and create a review.
Deploy on Laravel Cloud
- Push this repo to GitHub.
- In Laravel Cloud: New application → import the repo.
- Attach Postgres and object storage.
- Set env:
APP_NAME=ReviseMyFILESYSTEM_DISK/REVISEMY_DISKto the Cloud object storage diskAPP_URLto yourhttps://….laravel.cloudURL- Queue worker enabled (
QUEUE_CONNECTION) - Optional
OPENAI_API_KEYfor vision second opinion
- Deploy. Run migrations from Cloud commands:
php artisan migrate --force - Reply to the contest with your
https://….laravel.cloudhomepage URL.
Self-host
Point MCP at your own origin. Use S3-compatible storage for screenshots in production. Rate limits and 7-day review expiry are built in.
Stack
- Laravel 13
- Livewire 4 + Flux UI
- Laravel MCP (web endpoint)
- Laravel Sanctum try tokens
- Postgres / SQLite + object storage
Docs
- CONNECTORS.md — Cursor / Claude / VS Code / ChatGPT packaging
- SECOND-OPINION.md — second opinion, agent subagent findings, work packets
- DEPLOY.md — Laravel Cloud contest deploy
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。