visual-review-workbench
Enables visual review of frontend UI variants by providing an MCP tool that opens, refreshes, verifies, captures, and closes review surfaces.
README
Visual Review Workbench
A portable Agent Plugins 1.0.0 package for reviewing real frontend UI variants.
The plugin combines:
- An Agent Skill that defines the visual-review workflow and comparison rules.
- One MCP tool that opens, refreshes, verifies, captures, and closes review surfaces.
- Dependency-free Node.js scripts that implement the review shell.
Package boundaries
This package contains only reusable review infrastructure. It intentionally excludes:
- Product source code and product-specific review routes.
- Screenshots, recordings, captured application state, and business fixtures.
- Local usernames, absolute machine paths, browser profiles, credentials, and environment files.
- Repository commits, worktrees, and generated review output.
Project-specific adapters are generated at runtime inside the project being reviewed. Generated review output belongs in a temporary or project-selected directory, never in this plugin package.
Components
visual-review-workbench/
├── plugin.json
├── mcp.json
├── skills/
│ └── review-frontend-variants/
│ ├── SKILL.md
│ ├── references/
│ └── scripts/
├── server/
│ └── index.mjs
└── scripts/
The MCP server exposes one tool: visual_review_workbench.
Its action field supports open, refresh, verify, capture, and close. Route scaffolding, fixture capture, measurement, and recording remain internal scripts directed by the skill instead of expanding the public tool surface.
Browser-backed commands require Playwright to be resolvable from the target project or plugin runtime. Static generation and serving use only Node.js built-ins.
Validate
npm test
To add organization-specific deny terms to the portability audit:
PORTABILITY_DENY_TERMS="internal-product,example-user" npm run check:portable
Runtime paths
The package never embeds installation paths. The client supplies PLUGIN_ROOT and PLUGIN_DATA according to the Agent Plugins specification. MCP subprocess state and server logs use PLUGIN_DATA; target-project paths are accepted only as runtime tool arguments.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
Neon Database
MCP server for interacting with Neon Management API and databases
E2B
Using MCP to run code via e2b.
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.