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 skills add 0xmariowu/Autosearch --skill fetch-playwrightgit clone --depth 1 https://github.com/0xmariowu/AutosearchWrote 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/0xmariowu/autosearch/fetch-playwright)<a href="https://agentmods.dev/skills/0xmariowu/autosearch/fetch-playwright"><img src="https://agentmods.dev/badge/skills/0xmariowu/autosearch/fetch-playwright/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/skills/0xmariowu/autosearch/fetch-playwright"><img src="https://agentmods.dev/badge/skills/0xmariowu/autosearch/fetch-playwright.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 | $0.00071 | $0.01075 |
| Opus 5.5 | $0.00028 | $0.00430 |
| Sonnet 5.5 | $0.00014 | $0.00215 |
| Haiku 4.5 | $0.00007 | $0.00108 |
Grade A, and why
fetch-playwright 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 6d 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 — 89 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Drive a real Chromium browser from the runtime AI via Microsoft's official @playwright/mcp server. Use this when fetch-jina (cheapest) and fetch-crawl4ai (depth + basic JS) still cannot reach the content — typically because the page requires clicking, typing, scrolling, waiting for an event, or capturing a screenshot.
Architecture (why there is no Python file in this skill)
This skill is documentation-only. The runtime AI calls playwright-mcp tools directly through its MCP client — autosearch does not wrap them. The skill exists so the runtime AI knows:
- When to reach for Playwright (vs. fetch-jina / fetch-crawl4ai).
- How to install
@playwright/mcp. - Which MCP tools to call for common web-research patterns.
Install
Register with your MCP client. For Claude Code, prefer claude mcp add (writes to ~/.claude.json); for project-scoped config use <project>/.mcp.json. Cursor reads ~/.cursor/mcp.json; Zed reads <project>/.zed/mcp.json. Schema for any of them:
{
"mcpServers": {
"playwright": {
"command": "npx",
"args": ["@playwright/mcp@latest"]
}
}
}
On first run, npx will download @playwright/mcp and Playwright will install a Chromium bundle. No API keys. Works offline for local pages.
Common Tool Calls (as used by the runtime AI)
Tool names below come from @playwright/mcp. Exact schemas are served by the MCP server itself; the runtime AI enumerates them via the MCP client.
browser_navigate(url)— go to a URL.browser_snapshot()— accessibility tree of the current page (cheap text form, no screenshot).browser_click(ref)/browser_type(ref, text)/browser_press_key(key)— interact with elements from the snapshot.browser_wait_for(text | time | element)— wait for an event or selector.browser_take_screenshot()— capture the current page as an image.browser_evaluate(script)— run custom JS (use sparingly).browser_network_requests()/browser_console_messages()— inspection.browser_tabs(...)— open, list, select, close tabs.browser_close()— release the browser.
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.
- 6d ago First seen · 89 lines · 71 tokens per session scan A 59daa37b72be
fetch-playwright is a skill published in the GitHub repository 0xmariowu/Autosearch (50 stars, last pushed 7d ago), licensed MIT. It adds 71 tokens to every session and 1,075 once invoked, about $0.0003 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 2026-10-11.
Other skills, from other repositories
find-my-bookings-and-balances
Uses your own email and browser connectors together to find booking confirmations and points balances, then saves them to Award Travel Finder. Use when the user asks to find, sync or set up all their trips and balances and an email connector (Gmail, Outlook, Apple Mail) or browser control (Claude in Chrome, computer…
scrape-strategy
Guide mode selection when site scraping is weak, blocked, or low quality. Use when the user asks why scraping failed or how Primr should proceed on a protected site.
captchasonic-solve
Solve a CAPTCHA using the CaptchaSonic MCP server. Supports all types — PopularCaptcha, reCAPTCHA, Geetest, OCR, AWS WAF, TikTok, Binance, Cloudflare Turnstile, and more. Provide an image (path, URL, or base64) and captcha type.
captchasonic
Solve CAPTCHAs using the CaptchaSonic AI API. Supports reCAPTCHA, Cloudflare Turnstile, Geetest, AWS WAF, TikTok, Binance, OCR, and more. Provides MCP server tools for health checks, balance queries, and captcha solving from any AI agent.
captchasonic-setup
Set up the CaptchaSonic MCP server for the first time. Installs the server, configures environment variables, and verifies connectivity. Use when a user wants to start using CaptchaSonic with their AI agent.
playwright-cli
Automates browser interactions for web testing, form filling, screenshots, and data extraction. Use when the user needs to navigate websites, interact with web pages, fill forms, take screenshots, test web applications, or extract information from web pages.