Skip to content
Agentmart
Submit a tool

UI Debugger MCP

Autonomous UI debugging MCP server: an agent drives the browser/desktop, reports bugs + UX issues.

Connector by developerz-ai · GitHub account verified by the MCP Registry

IndexedMITLocal (stdio)★ 2
Write a review

Install

Works with Claude Code, Codex, VS Code, Cursor, Windsurf, Claude Desktop. Read the repository before you give any tool access to your data.

npx agentmarket install connector/developerz-ai-ui-debugger-mcp

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Autonomous UI debugging MCP server: an agent drives the browser/desktop, reports bugs + UX issues.

Reviews

Write a review

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this server can do and what it asks for, from its tool list and install config. Checked Oct 25, 2026.

Tools
2 read · 3 write · 0 delete
Secrets
1 (API keys, tokens or passwords)
Install scripts
Not scanned yet
Runs as
Node.js package

Tools by what they do 5

Read from the source code; the running server may list different tools.

Read 2 · Look things up without changing anything.

  • describe
  • get_findings

Write 3 · Create, change, send or run things.

  • end_session
  • send_message
  • start_debug

Secrets and settings it asks for

Secrets: values you paste in that let it act as you.

  • OPENAI_API_KEY

Other settings (environment variables or headers).

  • OPENAI_BASE_URL

Tools (5)

Read from the source code; the running server may differ. Descriptions are the server’s own words; hidden characters are shown as [U+…].

  • describeDescribe debug targetsread-only

    List the configured targets for this project (name, adapter, managed|attach, operational — whether the adapter kind is wired in this build and supported on the host platform; a wiring/capability fact, NOT a promise the target is up or that start_debug will succeed), web url/headless, any named auth personas, and the target's standing notes — what is expected of the app) plus the resolved per-role models and workspace. Call this first to pick a valid target — and a valid `as` persona — for start_debug. Pass target to narrow to one. Also reports `session`: the run this project currently holds, s…

  • end_sessionEnd the debug run

    Close the active run: abort the loop, release the target (managed Chrome is stopped; an attached browser is only disconnected), and free the project lock so a new start_debug can run. The last get_findings snapshot remains on disk.

  • get_findingsGet run findingsread-only

    Poll the run: status (running|passed|failed) plus structured findings — the step trail, functional bugs, visual/UX feedback, summary, and evidence paths (screenshots, logs). Pass wait (ms) to long-poll until the run settles; pass fields to return only some keys. A run that auto-ended (wall-clock timeout or client disconnect) stays readable under its id until end_session or the next start_debug.

  • send_messageMessage the running run

    Talk to the small driver agent mid-run: add work, redirect it, or answer a question. The message is folded into the conversation before the next step — the run keeps going, no restart. Use get_findings to see the effect.

  • start_debugStart a debug run

    Open a debug session: hand the small driver agent a goal for a configured target (e.g. "web"). It drives the UI autonomously and gathers findings (functional bugs + visual feedback). One run per project at a time. Returns { session_id }; then poll get_findings, steer with send_message, and close with end_session. A run is already active? The error names its session_id and how to read, close or replace it — and describe reports that id too.

Signals

Stars
2
Forks
0
Open issues
14
Weekly downloads
28
Last commit
Oct 23, 2026 (5 days ago)
Last release
Sep 5, 2026 (1 month ago)
Last package publish
Sep 5, 2026 (1 month ago)

From GitHub (fetched Oct 24, 2026) and the package registry.

Details

Registry name
io.github.developerz-ai/ui-debugger-mcp
Version
1.9.0
Auth
API key
Packages
npm @developerz.ai/ui-debugger-mcp
License
MIT
Repository
github.com/developerz-ai/ui-debugger-mcp
Website
github.com/developerz-ai/ui-debugger-mcp#readme
First seen
Oct 24, 2026
Last checked
Oct 25, 2026
Source
Official MCP Registrygithub-graphql
Cite this page

UI Debugger MCP. Agent Mart. https://agentmart.dev/connectors/developerz-ai-ui-debugger-mcp (last updated 2026-10-25)