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.
npx agentmods add skills/pandysp/claude-plugins/drive-browsernpx skills add pandysp/claude-plugins --skill drive-browsergit clone --depth 1 https://github.com/pandysp/claude-pluginsWrote 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/skills/pandysp/claude-plugins/drive-browser)<a href="https://agentmods.dev/skills/pandysp/claude-plugins/drive-browser"><img src="https://agentmods.dev/badge/skills/pandysp/claude-plugins/drive-browser.svg" alt="Measured on agentmods" 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 | $0.00172 | $0.01868 |
| Opus 5 | $0.00086 | $0.00934 |
| Sonnet 5 | $0.00034 | $0.00374 |
| Haiku 4.5 | $0.00017 | $0.00187 |
Grade A, and why
drive-browser 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 4d 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 — 120 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Drive browser with Playwright
Playwright is a superset of every scriptable browser-automation path: precise scripting AND a human-like vision loop, on a fresh browser OR the user's real logged-in one. The trap is following the API's grain. Reflexively reaching for brittle CSS, or jumping straight to vision, when the right tool sits in between. Pick by the target, not by habit.
Decide: two independent choices
How you find and act on elements:
- Locators (default).
getByRole/getByLabel/getByTextfirst: they ride the accessibility tree, auto-wait, and re-resolve as the DOM churns. CSS (page.locator('input[name="x"]')) only when a locator can't express the target. This handles your own app and most real sites. - Vision: last resort, only when the structure is genuinely opaque (custom canvas, a widget the a11y tree doesn't expose, selectors that won't resolve): screenshot → read it → click coordinates.
Where the page runs:
- Fresh launch: isolated, deterministic, can't touch the user's session. Tests, experiments, scraping.
- Attach to the real browser: the user's cookies/session, when the task needs their login (read Safety).
The axes are orthogonal: you can locator-debug the real attached session, or run vision on a fresh launch.
Locator/selector mode (precise)
Prefer resilient locators over raw CSS. page.getByRole('button', { name: 'Save' }),
getByLabel('Email'): they auto-wait and survive DOM churn (the right default for
slow SPAs). Drop to page.locator(css) / page.fill(css, v) only when a locator
can't express the target. Then page.evaluate(fn) for exact DOM reads, web-first
waits (expect(locator).toBeVisible(), waitForSelector, waitForURL), and
page.emulateMedia({ reducedMotion: 'reduce' }) for CSS-media features. Fast,
exact, reproducible. Use it to reproduce a bug, check a streaming caret, test
prefers-reduced-motion.
Vision mode (last resort, for genuinely opaque UI)
Drive by sight only when locators can't reach the controls. It's an iterative
loop with you as the eyes: page.screenshot() → read the PNG → decide
coordinates → page.mouse.click(x, y) → page.keyboard.type(text) →
re-screenshot to confirm. Don't write it as one blind script with guessed
coordinates.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 4d ago First seen · 120 lines · 172 tokens per session scan A 1de4109b4d3b
drive-browser is a skill published in the GitHub repository pandysp/claude-plugins (5 stars, last pushed 13d ago), licensed MIT. It adds 172 tokens to every session and 1,868 once invoked, about $0.0009 per session on Opus 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 2026-08-31.
Other skills, from other repositories
use-agent-browser-for-airi
Test AIRI display-model imports with agent-browser across stage-tamagotchi Electron, stage-web, and stage-pocket mobile web layouts. Use when uploading and verifying contributor-supplied Live2D ZIP, VRM, or MMD ZIP/PMX/PMD files through AIRI's model selector, including onboarding bypass, format-specific import…
opencli-sitemap-author
Use when creating or maintaining OpenCLI site sitemaps: agent-facing navigation, page-state, action, workflow, API-reference, pitfall, and fallback knowledge for a website. Use after browser exploration discovers durable site context, when a sitemap is stale, or when promoting local site knowledge into the repo.
debug-optimize-lcp
Guides debugging and optimizing Largest Contentful Paint (LCP) using Chrome DevTools MCP tools. Use this skill whenever the user asks about LCP performance, slow page loads, Core Web Vitals optimization, or wants to understand why their page's main content takes too long to appear. Also use when the user mentions…
webapp-testing
Toolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing browser screenshots, and viewing browser logs.
interactive-login
How to complete browser/interactive logins (aws / gh / glab / gcloud). The platform backgrounds the login poller so it survives the human's browser round-trip — and when that does NOT work.
pinchtab-mcp
Use this skill when a task requires browser automation through PinchTab's MCP server connected to a remote browser instance. Covers navigation, element interaction, data extraction, form filling, multi-step flows, and session management via MCP tools.