Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/rm2thaddeus/Pixel_DetectiveWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/rules/rm2thaddeus/pixel_detective/mcp-browser-tools-setup)<a href="https://agentmods.dev/rules/rm2thaddeus/pixel_detective/mcp-browser-tools-setup"><img src="https://agentmods.dev/badge/rules/rm2thaddeus/pixel_detective/mcp-browser-tools-setup/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/rules/rm2thaddeus/pixel_detective/mcp-browser-tools-setup"><img src="https://agentmods.dev/badge/rules/rm2thaddeus/pixel_detective/mcp-browser-tools-setup.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00008 | $0.01460 |
| Opus 5.5 | $0.00003 | $0.00584 |
| Sonnet 5.5 | $0.00002 | $0.00292 |
| Haiku 4.5 | $0.00001 | $0.00146 |
Grade A, and why
mcp-browser-tools-setup scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl http://localhost:3025/health How it starts
The opening of the file, as written. The whole thing — 196 lines — stays where its author put it; the contents beside it link to each section on GitHub.
MCP Browser Tools Setup & Troubleshooting Protocol
🚨 CRITICAL: Browser Tools MCP Requires 3-Component Setup
Sprint 10 revealed that Browser Tools MCP is COMPLEX and fails silently without proper setup. ALWAYS verify all components.
✅ MANDATORY 3-COMPONENT VERIFICATION:
1. Node.js Version Check (CRITICAL)
# ✅ MUST be Node.js v18+ (requires native fetch API)
node --version # Must show v18.0.0 or higher
# ❌ If version is <18:
nvm install 20
nvm use 20
nvm alias default 20
2. Browser Tools Server (Port 3025)
# ✅ Start the middleware server
npx @agentdeskai/browser-tools-server@latest
# ✅ Verify server is running
netstat -an | findstr 3025 # Windows
lsof -i :3025 # Mac/Linux
# ✅ Should see output like:
# TCP 0.0.0.0:3025 0.0.0.0:0 LISTENING
3. Chrome Extension Installation
# ✅ Clone and install extension
git clone https://github.com/AgentDeskAI/browser-tools-mcp.git
cd browser-tools-mcp
Chrome Extension Steps:
- Open Chrome →
chrome://extensions/ - Enable "Developer mode" (top right toggle)
- Click "Load unpacked"
- Select folder:
browser-tools-mcp/chrome-extension - Verify "BrowserToolsMCP" appears in extensions list
🔍 VERIFICATION PROTOCOL:
Test MCP Connection
# ✅ Test 1: Basic MCP connectivity
mcp_browser-tools_wipeLogs
# Expected: "All logs cleared successfully"
# ✅ Test 2: Server communication
mcp_browser-tools_getConsoleLogs
# Expected: [] (empty array) or actual logs
# ✅ Test 3: Full functionality (requires extension)
mcp_browser-tools_takeScreenshot
# Expected: Screenshot data or specific error message
Extension Connection Check
- Open any webpage (e.g., https://example.com)
- Open Chrome DevTools (F12)
- Look for "BrowserTools" tab in DevTools
- Verify connection status shows "Connected"
❌ COMMON ERROR PATTERNS & FIXES:
Error: "Failed to discover browser connector server"
# ✅ FIX: Start Browser Tools Server
npx @agentdeskai/browser-tools-server@latest
# ✅ Verify server is accessible
curl http://localhost:3025/health
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 196 lines · 8 tokens per session scan A 9b97df5c9aca
mcp-browser-tools-setup is a cursor rule published in the GitHub repository rm2thaddeus/Pixel_Detective (21 stars, last pushed 8mo ago), licensed MIT. It adds 8 tokens to every session and 1,460 once invoked, about $0.0000 per session on Opus 5.5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other cursor rules, from other repositories
browser-diagnosis
Reproducing and diagnosing App Builder in the browser — SDV, Playwright scratch drivers, selection menu pitfalls, PUT counting. Use when debugging UI, selection, or interaction.
autofix
Please use the Playwright MCP to open a browser to the URL I provide, check for any errors, and then automatically fix those errors in my codebase, then continue to iterate until all the errors are gone.
mcp-inspector-debugging
Debugging and verifying MCP servers using the MCP Inspector UI with Playwright automation.
native_browser_debugging
Use native browser for UI verification and debugging; check console for errors. Load when investigating or fixing UI, OAuth, or frontend behavior.
can-decode
Check if a video can be decoded by the browser using Mediabunny.
ponytail
Ponytail, lazy senior dev mode. Always pick the simplest solution that works.