Mmcp.market

Thinkrun MCP server

by dundas·io.github.dundas/thinkrun·v0.3.8
No reviews yet

Browser control + session recording MCP for AI coding agents (Claude Code, Cursor) — real Chrome.

A92/100grade A
Write a reviewInstall
What users say
No reviews yet
Be the first
Safety scan
A92/100

full report

Adoption
Growing

2 stars44 downloads/wk

Reviews

Write one

Nobody has reviewed Thinkrun yet.

If you have run it, two minutes of your experience saves the next person an afternoon.

Thinkrun tools (54, 7 write)

write = sends, deletes, buys or posts

Read from the package source without running it. The installed server may list more.

  • clear_logs

    Clear console and network logs for the session.

  • click

    Click an element on the page using a CSS selector.

  • click_at

    Click at viewport-relative CSS pixel coordinates. Local mode only.

  • console_messages

    Get browser console messages (log, warn, error) from the current session.

  • evaluatewrite action

    Execute JavaScript in the browser page context. The script runs in the page,

  • extract

    Extract content from the page. Types: "text" for visible text, "html" for raw HTML,

  • fill

    Fill a form field with a value (replaces existing content).

  • focus

    Bring the currently bound local browser tab/window to the foreground without rebinding the default session. Local mode only.

  • get_dialog

    Check for pending browser dialogs (alert, confirm, prompt) and view dialog history.

  • get_html

    Return the current page HTML for the active session. Large documents are truncated explicitly in the response.

  • get_title

    Return the current page title for the active session.

  • get_url

    Return the current page URL for the active session.

  • go_back

    Navigate back to the previous page in browser history.

  • go_forward

    Navigate forward to the next page in browser history.

  • handle_dialog

    Accept or dismiss a pending browser dialog (alert, confirm, prompt).

  • hover

    Hover the mouse over an element. Useful for triggering tooltips or dropdown menus.

  • local_action_cancelwrite action

    Cancel a running local action. Local mode only.

  • local_action_runwrite action

    Run a direct native-host action against the currently attached local tab.

  • local_action_status

    Check the status of a previously started local action. Local mode only.

  • local_diagnostics

    Return structured local continuity, ownership, reclaim, and bridge recovery diagnostics. Local mode only.

  • local_reset_connection

    Trigger the bounded native-host reset-connection recovery path used for false local circuit-breaker trips. Local mode only.

  • navigate

    Navigate the browser to a URL. Waits for page load before returning.

  • network_requests

    Get network requests made by the browser during this session.

  • page_cache_html

    Stateless: fetch rendered HTML for a URL without a browser session (no sessionId).

  • page_cache_screenshot

    Stateless: capture a PNG for a URL without a session. Returns JSON with artifactUrl when Accept is application/json.

  • page_cache_text

    Stateless: extract plain text for a URL without a browser session. POST /api/cache/text.

  • press_key

    Press a keyboard key. Common keys: Enter, Tab, Escape, ArrowDown, ArrowUp, Backspace.

  • screenshot

    Capture a screenshot of the current page. Returns an image plus any available artifact metadata.

  • scroll

    Scroll the page or a specific element. Positive y scrolls down, negative scrolls up.

  • select_option

    Select an option from a <select> dropdown element.

  • session_artifacts

    List artifacts (screenshots, recording) for a session.

  • session_createwrite action

    Create a new browser session. Returns a sessionId to use with all other tools.

  • session_deletewrite action

    Terminate a browser session and clean up resources. Artifacts are preserved.

  • session_list

    List all active browser sessions for this user.

  • session_release

    Release the current local tab ownership without closing the tab. Clears the default session when releasing the bound local session. Local mode only.

  • session_status

    Get the current status of a browser session including URL, title, and action count.

  • session_use

    Set the default session for subsequent tool calls so you do not need to pass sessionId every time.

  • session_wait_ready

    Poll session status until the session is ready (running/active) or timeout.

  • set_auditing

    Enable or disable local audit mode for the current local tab-backed session.

  • set_mode

    Switch the MCP server between local (Chrome extension) and cloud (ThinkRun API) modes at runtime.

  • sleep

    Pause for a bounded duration in milliseconds (max …).

  • snapshot

    Get an accessibility tree snapshot of the current page. This is a lightweight

  • tab_attach

    Attach to an existing local Chrome tab by tab ID and make it the default session. Local mode only.

  • tab_close

    Close a browser tab by ID. If no tabId is provided, closes the current tab. Local mode only.

  • tab_list

    List all open browser tabs. Local mode only.

  • tab_new

    Open a new browser tab, optionally navigating to a URL. Local mode only.

  • tab_switch

    Switch the active local Chrome tab without rebinding the current default session. Local mode only.

  • task_createwrite action

    Create an AI task plan from natural language. The plan breaks your request into

  • task_executewrite action

    Execute a previously created task plan. Cloud mode only.

  • task_status

    Check the status of a task plan execution. Cloud mode only.

  • type_text

    Type text into an element (appends to existing text, triggers key events).

  • wait_for_element

    Wait for an element to appear on the page. Useful after navigation or clicking

  • wait_for_text

    Wait for specific text to appear on the page.

  • window_new

    Open a new local Chrome window and bind the resulting tab as the default session. Local mode only.

Public scan report

scanner v0.1.9 · 2026-09-26 · same rubric, same numbers if you re-run it

no findings
  • Code scan25 source files scanned25/25
  • –Live reliabilityno gateway calls yet and no remote to proben/a
  • –Tool poisoningtools not inspected (local package is not executed); not countedn/a
  • Auth qualitylocal package, no credentials required12/15
  • Maintenancelast push 5 days ago15/15
  • Maintainer identityregistry namespace matches repository owner; GitHub account older than a year8/10
Overall 92/100. Components that don't apply are left out of the denominator. Any critical finding is an F.RubricAppeal a findingJSON

What the publisher says

From the Thinkrun repository's README, as published. We do not edit it. Read it on GitHub

ThinkRun — Screen Recorder for AI Coding Agents

Stop screenshotting bugs and typing paragraphs to explain them. ThinkRun records any browser session — or lets your agent record what it built — and turns it into structured context your AI coding agent can act on: every click, the console, network requests, your voice narration, and screenshots, all time-synced. Not just a screenshot. Not just a video.

Works with Claude Code, Cursor, Cline, Windsurf, and any MCP client.

Install

npx @thinkrun/mcp

Or install the CLI globally:

npm install -g @thinkrun/cli

The CLI automatically installs the native host binary, which lets AI tools control Chrome.

Add to your MCP config

{
  "mcpServers": {
    "thinkrun": {
      "command": "npx",
      "args": ["@thinkrun/mcp"]
    }
  }
}

Config location:

  • Claude Code: ~/.claude/settings.json
  • Cursor: .cursor/mcp.json
  • Cline: VS Code settings → Cline MCP Servers

Then ask your agent: "navigate to localhost:3000 and screenshot the checkout page" — or record the bug yourself and hand the agent the result.

Why ThinkRun

Other tools give your agent eyes. ThinkRun gives it structured sight — context it can parse in one pass instead of you re-explaining the page.

Agent skills for Claude Code, Cursor, Codex & Gemini CLI

Six portable agent skills ship in this repo, mirrored for every major harness (.claude/skills/, .cursor/skills/, .codex/skills/, .gemini/skills/):

  • prove-it — the prove step of a software-factory workflow: named test targets in, a pass/fail table with a screenshot, console and network capture per target out, posted to the PR; any non-pass row sends the change back to build.
  • before-after — the ship step: two captures at the same viewport, one | Before | After | table for the PR, no public paste host.
  • web-browse — browse the web programmatically: drive a real or cloud browser to navigate, interact, extract, and screenshot.
  • ux-audit — walk through a product UI as a real user, screenshot every step, and produce a structured UX report with a prioritized fix list.
  • thinkbrowse-cli — drive the browser from shell scripts via the thinkrun CLI.
  • thinkbrowse-mcp — drive the browser through the ThinkRun MCP server's tools.

Install without cloning — the repo is served as a live MCP server that auto-updates on every commit:

{ "mcpServers": { "thinkrun-skills": { "url": "https://mcp.skillsovermcp.com/mcp/dundas/thinkrun" } } }

Trigger accuracy is measured, not claimed. Every skill ships a 20-query trigger-eval fixture, and evals/BENCHMARK.md publishes the results: with the six-skill catalog (v4, 2026-09-18), 120/120 on GPT-5.6-terra and Claude Fable 5 and 113/120 on Claude Sonnet 5 (its documented strict-MCP reading, not a crash); the earlier four-skill matrix scored 79–80/80 on 12 of 16 model configurations. Catalog-ablation runs show the browsing skills absorb each other's traffic when installed as a subset, and that prove-it fills a gap rather than duplicating a sibling (before-after folds into prove-it when absent). Reproduce with bun evals/run-eval.ts build.

Manual binary install

If you need to install the native host without the CLI, download the binary for your platform from the latest release and run:

chmod +x thinkbrowse-host-* && ./thinkbrowse-host-* --install

Shortened. The full README is on GitHub.

Nothing above is checked by us. What we check is on the safety report.

Install directly

Runs npx -y @thinkrun/mcp on your machine. Read the scan report first; the gateway never runs local packages.

claude mcp add thinkrun -- npx -y @thinkrun/mcp
Add to Cursor

Thinkrun: common questions

Is Thinkrun MCP server safe?
Yes, by our scan: it is graded A (92/100). Read the Thinkrun safety report
How do I install Thinkrun?
It runs on your machine. Copy the Claude Code, Claude Desktop or Cursor config from the install section.
Does Thinkrun need an API key?
Not as far as the registry entry and our scan can tell: no credentials are declared or required.
Is Thinkrun maintained?
The last commit was 6 days ago (2026-09-21). The latest release is v0.3.8.
What can I use instead of Thinkrun?
Servers from other publishers that do the same job: Browser Use MCP server, Zenrows MCP server and peek MCP server. Compare all Thinkrun alternatives.

Alternatives to Thinkrun

Same job from other publishers: the closest match first, then the best rated.

All Thinkrun alternatives →
  • Browser Use
    Control a real Chrome browser to complete any task: fill forms, extract data, book flights.
    A
  • Zenrows
    Zenrows MCP server — Fetch, Extract, Batch, and Browser Sessions for AI coding assistants
    A
  • peek
    Exposes recorded browser sessions (rrweb DOM, console, network) to AI coding agents, fully local
    B
  • Safari
    Native Safari browser automation for AI agents — 98 tools, zero Chrome overhead.
    C
  • Agent360 Browser MCP
    Drive your real, logged-in Chrome from any AI agent - 40 tools. Needs our Chrome extension too.
    A
Thinkrun vs Browser UseThinkrun vs ZenrowsThinkrun vs peek

More from dundas →