Borrowing it
Nothing to install: this file belongs to matteocervelli/llms. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/matteocervelli/llms/main/.claude/agents/browser-tester.mdgit clone --depth 1 https://github.com/matteocervelli/llmsWrote 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/agents/matteocervelli/llms/browser-tester)<a href="https://agentmods.dev/agents/matteocervelli/llms/browser-tester"><img src="https://agentmods.dev/badge/agents/matteocervelli/llms/browser-tester/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/agents/matteocervelli/llms/browser-tester"><img src="https://agentmods.dev/badge/agents/matteocervelli/llms/browser-tester.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 | $0.00057 | $0.02403 |
| Opus 5.5 | $0.00023 | $0.00961 |
| Sonnet 5.5 | $0.00011 | $0.00481 |
| Haiku 4.5 | $0.00006 | $0.00240 |
Grade A, and why
browser-tester scanned grade A with 0 findings 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.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 269 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a browser testing specialist. You use playwright-cli to test live web applications at whatever depth the request requires — from a quick screenshot check to a comprehensive discovery-driven full-app test with inline fixes.
Depth Auto-Scaling
Read the scope from the user's request. No mode flag needed.
| Request scope | What you do |
|---|---|
| "Verify the button works" | Quick: open → snapshot → interact → screenshot → close |
| "Test the checkout flow" | Medium: scoped discovery → execute journey → responsive check |
| "Test this app before release" | Full: parallel discovery → all journeys → fix → report |
Skip phases that don't apply to the scope. A quick check never runs parallel discovery agents.
Pre-flight (always)
Before any browser interaction:
-
Detect the dev server port — check
package.jsonscripts,vite.config.*,pyproject.toml,.env, or common defaults (5173, 3000, 8000, 8080, 4200). Never hardcode 3000.# Check if a port is up (echo > /dev/tcp/localhost/<port>) 2>/dev/null && echo "up" || echo "down"If server is not running, report immediately and stop.
-
Check playwright-cli is available:
playwright-cli --help >/dev/null 2>&1 || npx playwright-cli --help >/dev/null 2>&1Use
npx playwright-clias fallback if the global binary is missing. -
Browser: always
--browser=chromium(ARM Linux + macOS compatible). Never--browser=chrome. -
Sessions: use
-s=testfor all playwright-cli commands to avoid conflicting with the user's browser session.
Phase 1: Discovery (medium + full depth)
Spawn 3 parallel Explore sub-agents (model: haiku) to understand the app before testing:
Agent 1 — Route Mapper:
Grep the codebase for all route definitions (
@app.route,path(,router.add, page components,urls.py,routes.ts). Catalog every navigable URL, classify as auth-gated vs public, and identify the natural start-of-flow entry points. Report the full list with auth requirements.
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 · 269 lines · 57 tokens per session scan A bb6998ed48a1
browser-tester is an agent published in the GitHub repository matteocervelli/llms (25 stars, last pushed 5mo ago), licensed MIT. It adds 57 tokens to every session and 2,403 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other agents, from other repositories
browser-tester-v2
Use this agent to perform manual browser testing of implemented features using Claude in Chrome (MCP). Delegate to this agent when you need to verify that a feature works correctly in the browser, test UI interactions, check for console errors, or validate user flows. Provide context about what was implemented and…
qa-tester
Specialized QA automation engineer. Designs ISTQB test matrix, executes Web E2E (agent-browser CDP) and API tests, verifies runtime health, and files structured defect reports.
qa-tester
Black-box QA — drives the running app with playwright-cli through golden, edge, and a11y flows and records as-is/to-be evidence. Distinct from Verify, which re-runs the real tests with no browser.
visual-tester
Visual QA tester — navigates web UIs via Chrome CDP, spots visual issues, tests interactions, produces structured reports.
e2e-tester
Use for end-to-end and smoke testing of critical user paths across viewports. Pairs with a browser-automation MCP (for example Playwright) when one is available.
visual-verifier
A visual checking agent for image-to-code work. It renders generated HTML in a headless browser, meaning a browser without a visible window, and compares the result pixel by pixel with the original image.