Apify Agent Skills — Setup Guide¶
Source: apify/agent-skills (2,232⭐) Category: Web Scraping / Automation Language: Python
Production-grade web scraping and automation skills for AI coding agents. Drop these into any agent and get expert access to the Apify platform — scrape any site, build new Actors, Actorize existing code, generate output schemas, and integrate the Apify API. 30,000+ pre-built Actors covering social media, search, maps, real estate, reviews, and more.
Installation¶
# Install all Apify agent skills
npx skills add apify/agent-skills
Prerequisites¶
| Requirement | Details |
|---|---|
| Apify account | console.apify.com — free tier available ($5/month credit) |
| Apify API token | From Apify Console → Settings → Integrations → API token |
| MCP server (optional) | npx @apify/mcp for MCP-native tool access |
| Hermes Agent | Any version — skills load as instructions; MCP integration available |
Set environment variables:
export APIFY_TOKEN="apify_api_..."
# Optional: MCP config in ~/.hermes/config.yaml
Key Capabilities¶
Available Skills¶
| Skill | What It Does |
|---|---|
| Scrape any site | Built-in Actor selection across 130+ curated Actors (Instagram, Facebook, TikTok, YouTube, X, LinkedIn, Google Maps, Reddit, Yelp, Airbnb) plus automatic fallback to 30,000+ Store Actors |
| Build new Actors | Generate, debug, and deploy serverless Actors in JS, TS, or Python with official SDK patterns |
| Actorize existing code | Wrap any script, library, or CLI tool as a runnable Actor with input/output handling |
| Generate output schemas | Auto-derive dataset_schema.json, output_schema.json, key_value_store_schema.json from Actor source |
| Integrate Apify API | Call Actors programmatically from JS/TS or Python via apify-client or REST API |
Supported Platforms (Curated Actors)¶
Social: Instagram, Facebook, TikTok, YouTube, X/Twitter, LinkedIn, Reddit Maps: Google Maps, Apple Maps Commerce: Amazon, eBay, Etsy, Alibaba Reviews: Yelp, TripAdvisor, Google Reviews Real Estate: Zillow, Airbnb, Realtor.com Search: Google Search, Bing, DuckDuckGo
CorpusIQ Use Cases¶
| Use Case | How |
|---|---|
| Competitive intel | Scrape competitor pricing, features, and reviews across platforms |
| Lead generation | Extract business listings from Google Maps, Yelp, LinkedIn |
| Social monitoring | Track brand mentions, follower growth, and content performance across platforms |
| Market research | Gather industry data, pricing trends, and customer sentiment at scale |
| Content discovery | Find trending content, hashtags, and creators in target niches |
Troubleshooting¶
| Issue | Fix |
|---|---|
| APIFY_TOKEN not recognized | Export as env var or add to ~/.hermes/config.yaml under env: |
| Actor timeout | Increase timeout: maxResults and maxConcurrency in Actor input |
| Rate limited by target site | Use Apify's built-in proxy rotation (residential + datacenter proxies) |
| MCP connection refused | Start MCP server: npx @apify/mcp then configure in Hermes MCP settings |
Verification¶
# Verify token works
curl -s "https://api.apify.com/v2/users/me?token=$APIFY_TOKEN" | python3 -m json.tool | head -5
# Verify skill installed
npx skills list | grep apify
# Quick test — list available Actors
curl -s "https://api.apify.com/v2/acts?token=$APIFY_TOKEN&limit=5" | python3 -c "import json,sys; d=json.load(sys.stdin); print([a['name'] for a in d.get('data',{}).get('items',[])])"
← Skills Catalog | Discovery Page → Powered by CorpusIQ