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/agentcomputerai/torch/bookingnpx skills add AgentComputerAI/torch --skill bookinggit clone --depth 1 https://github.com/AgentComputerAI/torchWhat 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 | $0.00118 | $0.02025 |
| Opus 5 | $0.00059 | $0.01012 |
| Sonnet 5 | $0.00024 | $0.00405 |
| Haiku 4.5 | $0.00012 | $0.00202 |
Grade A, and why
booking 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 yesterday.
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.
description: Proven scraping playbook for booking.com searchresults.html pages. CloudFront-fronted JS challenge blocks bare curl (HTTP 202 with a script-only interstitial), but a real Chrome session via the real Chrome d How it starts
The opening of the file, as written. The whole thing — 145 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Booking.com (booking.com)
Hotel/stay search results on
/searchresults.html?ss=<destination>. Real Chrome via127.0.0.1:9222clears the CloudFront JS challenge on first nav. Cards are in the DOM (not the initial HTML), pagination is click-to-load-more, and prices are absent unless dates are in the URL.
Detection
| Signal | Value |
|---|---|
| CDN | CloudFront (server: CloudFront, x-amz-cf-*) |
| Framework | Custom edge (MFE shell: web-shell-header-mfe, web-shell-footer-mfe) |
| Anti-bot | CloudFront JS interstitial on bare curl (HTTP 202, ~8 KB, reportChallengeError(...) in script) |
| Auth | Not required for search results |
| robots.txt | Disallows many deep paths; /searchresults.html is allowed |
Architecture
/searchresults.html?ss=<q>returns a tiny shell + JS that hydrates into a React-ish SPA.- The first paint server-renders 1 placeholder card; the rest are injected client-side.
- Initial batch is 25 cards; more load via a "Load more results" button (batches of ~25).
- A sign-in modal sometimes pops up — dismiss with
[aria-label="Dismiss sign-in info."]. - Each card has a stable
data-testid="property-card"wrapper with sub-testids fortitle,title-link,distance,review-score,price-and-discounted-price,price-for-x-nights,recommended-units,address-link.
Strategy used
- Phase 0 (curl): ❌ HTTP 202 challenge page, no data.
- Phase 1 (framework JSON): ❌ no
__NEXT_DATA__, no global payload blob worth parsing. - Phase 2 (browser): ✅
puppeteer.connect({ browserURL: "http://127.0.0.1:9222" })— zero challenges, cards render immediately. No stealth plugin required.
Stealth config that works
import puppeteer from "puppeteer-core";
const browser = await puppeteer.connect({ browserURL: "http://127.0.0.1:9222" });
const page = await browser.newPage();
await page.setViewport({ width: 1440, height: 900 });
await page.goto(url, { waitUntil: "domcontentloaded", timeout: 60000 });
// IMPORTANT: disconnect, do not close
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.
- yesterday First seen · 145 lines · 118 tokens per session scan A ee75b936e7f8
booking is a skill published in the GitHub repository AgentComputerAI/torch (5 stars, last pushed 4mo ago), licensed MIT. It adds 118 tokens to every session and 2,025 once invoked, about $0.0006 per session on Opus 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 2026-08-31.
Other skills, from other repositories
tiny-web-crawler
Crawl from one or more starting web pages, fetch readable content, search within pages, follow relevant links, and stop when the requested information is found or a bounded limit is reached.
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.
browserstack
../../../engineering-team/playwright-pro/skills/browserstack/SKILL.md.
analyze-performance-traces
Analyze Chrome, Chromium, Electron, React DevTools, or Perfetto-compatible JSON traces and audit user-reported profiling findings without loading large artifacts into context; prove trigger-to-render/layout chains, separate measured facts from source inference, find exact code choke points, classify forced layout and…
feature-demo-recording
Record a demo video of a web feature from a real browser. Two modes -- a NARRATED film where measured voiceover drives the timeline (designed slides, subtitles, punch-in camera, rendered from an HTML timeline), and a SILENT evidence clip for a PR or a QA pass. Use when the user asks to record a video, demo, or screen…
browser-recording
Record a browser flow as a video/GIF for evidence — animations, transitions, and multi-step interactions that a still screenshot cannot prove. Drives the project's own Playwright through a bundled runner, then converts to mp4 + GIF via ffmpeg. Use when the user asks to record a demo, capture a GIF or video of the UI…