Scrapfly Skills

Agent skills for web scraping, extraction, screenshots, crawling, browser automation, the scrapfly CLI, and the Scrapfly MCP toolset.

Compatible with Claude Code, Cursor, Cline, Codex, Windsurf, Gemini, GitHub Copilot, Roo, Goose, Antigravity, and 50+ other agents via the open skills CLI.

Install

Pick the flow that matches what you're shipping:

# Install every Scrapfly skill into every detected agent
npx skills add scrapfly/skills --all

# Pick specific skills (use the canonical `--skill <name>` form)
npx skills add scrapfly/skills --skill scrapfly-scraper
npx skills add scrapfly/skills --skill scrapfly-cli --skill scrapfly-agent-rules

# Target specific agents (claude-code, codex, cursor, gemini-cli, ...)
npx skills add scrapfly/skills --all -a claude-code -a cursor

# Install globally (~/<agent>/skills/) instead of in the current project
npx skills add scrapfly/skills --all --global

# Just list what's available without installing
npx skills add scrapfly/skills --list

scrapfly/skills is GitHub-shorthand for https://github.com/scrapfly/skills; the CLI clones it directly — no Scrapfly account or registry call required.

Skills

Skill	When to use
scrapfly-scraper	Python SDK agent calling scrapfly-sdk for scraping with proxy rotation, anti-bot bypass, and JS rendering
scrapfly-extraction	Extract structured data from HTML/markdown using LLM prompts, AI models, or templates
scrapfly-screenshot	Capture web page screenshots with JS rendering and anti-bot bypass
scrapfly-crawler	Crawl entire websites with link discovery, depth control, and data extraction
scrapfly-browser	Automate cloud browsers via Playwright with proxy rotation and fingerprinting
scrapfly-cli	Shell-tool agents (Claude Code, Codex, Aider) driving the scrapfly CLI binary instead of writing Python
scrapfly-agent-rules	Cross-tool golden rules for an LLM agent connected to the Scrapfly MCP server (stateless vs. browser, when to unblock, WebMCP)
scrapfly-sdk-typescript	Web scraping from TypeScript/JavaScript (Node.js, Deno, Bun) with the scrapfly-sdk package
scrapfly-sdk-go	Web scraping from Go with the go-scrapfly module and its built-in goquery selector
scrapfly-sdk-rust	Web scraping from Rust with the async scrapfly-sdk crate

Setup

For the Python SDK skills (scrapfly-scraper, scrapfly-extraction, scrapfly-screenshot, scrapfly-crawler, scrapfly-browser):

pip install scrapfly-sdk
export SCRAPFLY_API_KEY="scp-live-..."

For the other-language SDK skills, install the matching package:

npm install scrapfly-sdk                     # scrapfly-sdk-typescript
go get github.com/scrapfly/go-scrapfly       # scrapfly-sdk-go
cargo add scrapfly-sdk                       # scrapfly-sdk-rust
export SCRAPFLY_API_KEY="scp-live-..."

For the CLI skill (scrapfly-cli):

curl -fsSL https://scrapfly.io/scrapfly-cli/install | sh
export SCRAPFLY_API_KEY="scp-live-..."

For the agent-rules skill (scrapfly-agent-rules): nothing to install on the agent side — the rules assume your agent is connected to a Scrapfly MCP server. See agent-ai for reference ADK bootstraps in Python and Go.

Get your API key at scrapfly.io.

Updating & removing

# Pull the latest version of every installed Scrapfly skill
npx skills update

# Remove specific skills
npx skills remove scrapfly-scraper

# Remove every Scrapfly skill from a specific agent
npx skills remove --skill 'scrapfly-*' -a claude-code

Where skills land

npx skills symlinks the skills into each agent's conventional directory. A few examples:

Agent	Project install	Global install
Claude Code	.claude/skills/	~/.claude/skills/
Codex	.agents/skills/	~/.codex/skills/
Cursor	.agents/skills/	~/.cursor/skills/
Gemini CLI	.agents/skills/	~/.gemini/skills/
OpenCode	.agents/skills/	~/.config/opencode/skills/

Full agent matrix: see skills package README.