oxylabs/agent-skills

Oxylabs Agent Skills – Quick Overview

What it is: A collection of agent‑skill definitions that let large‑language‑model agents (e.g., Claude, other autonomous agents) talk directly to Oxylabs’ suite of web‑access services. Each skill packages the API contract, authentication details and usage tips for a specific product.

Skill	Core purpose	Typical use‑case
Proxies	Access residential, mobile, datacenter, ISP and dedicated proxy pools with geo‑targeting, IP rotation and sticky sessions.	An agent needs to fetch a URL while appearing to come from a particular country or device type.
Web Unblocker	AI‑enhanced proxy that automatically handles fingerprinting, JavaScript rendering and CAPTCHAs.	Retrieve a page that blocks bots or requires heavy client‑side rendering without launching a full browser.
Web Scraper API	Structured JSON extraction for >40 common targets (search engines, e‑commerce, etc.) with optional geo‑targeting.	An agent wants clean, ready‑to‑consume data (product listings, search results) instead of raw HTML.
Agent Browser	Remote, controllable browser sessions via CDP (Playwright/Puppeteer) backed by residential proxies.	Perform interactive actions – clicks, form fills, screenshots – that need a real browser environment.
Video Data	Metadata, subtitles, search, channel info and high‑bandwidth video/audio downloads from platforms like YouTube.	Gather video‑level insights or download media for downstream processing.

How it’s installed

• For Claude Code: add the plugin from the marketplace (/plugin marketplace add oxylabs/agent-skills).
• For any other agent framework: run npx skills add https://github.com/oxylabs/agent-skills.git.

Configuration All skills rely on environment variables stored in a .env file. Separate username/password pairs are provided for each product family (residential/mobile, datacenter/ISP, Web Unblocker, Web Scraper API, Agent Browser, Video Data). The Video Data skill also needs a dedicated high‑bandwidth endpoint variable.

Typical workflow

1. Pick the most specific skill that matches the task.
2. Issue a tiny test request to confirm authentication and expected response shape.
3. Scale up, respecting any rate‑limit or payload limits documented by Oxylabs.

Who it’s for Developers building autonomous agents that need reliable, location‑aware web access without writing low‑level HTTP or browser‑automation code themselves.

Further reading The README points to Oxylabs’ official agent skills documentation for deeper details on each skill’s parameters and response formats.