apify

Apify Skill

Dynamic research automation using Apify actors. Searches the Apify store to find the best actor for any research need, with a registry of validated actors for common tasks.

When to Activate

• User needs to scrape websites, job boards, or social media
• Research requires data that's not available via standard APIs
• Outreach campaigns need EDP data from external sources (job postings, company news, etc.)
• User mentions "scrape", "extract", "job postings", "LinkedIn search", or similar research needs

Core Philosophy

1. Dynamic Actor Selection

When a research need arises:

1. Check the actor registry for known good actors
2. If not found, search Apify store for appropriate actors
3. Prefer pay-per-result/event/usage pricing over flat subscription
4. Always ask user before building new research pipelines

2. Pay-Per-Result Preference

Always prefer actors with:

• Pay per result
• Pay per event
• Pay per usage

Avoid when possible:

• Flat monthly subscriptions
• High minimum costs

3. Build on Demand

When no suitable actor exists in the registry:

1. Search Apify store
2. Present options to user with pricing
3. Ask: "Should I add this to the registry for future use?"

---

Actor Registry

Known, validated actors for common research tasks. See modules/ACTOR_REGISTRY.md for full registry.

Quick Reference

Use Case	Actor	ID	Pricing
LinkedIn Jobs	fantastic-jobs/advanced-linkedin-job-search-api	vIGxjRrHqDTPuE6M4	Pay per result
YouTube Transcripts	karamelo/youtube-transcripts	-	Pay per result
People Database	apify/people-database	IoSHqwTR9YGhzccez	Pay per result

---

Usage

Run a Known Actor

# LinkedIn Jobs Search
uv run .claude/skills/apify/scripts/run_actor.py \
  --actor "vIGxjRrHqDTPuE6M4" \
  --input '{"searchQuery": "operations manager", "location": "United States", "maxResults": 100}' \
  --output jobs.json

# Or use the alias
uv run .claude/skills/apify/scripts/linkedin_jobs.py \
  --query "operations manager" \
  --location "United States" \
  --company-size "11-50,51-200" \
  --max-results 100

Search Apify Store

# Find actors for a research need
uv run .claude/skills/apify/scripts/search_store.py \
  --query "linkedin job postings" \
  --pricing "pay_per_result"

Build New Research Pipeline

When the system identifies a gap (like Tier 1 EDP needing job data):

1. Identify the need: "We need job posting data for agencies hiring ops roles"
2. Check registry: Is there a known actor for this?
3. Search store: If not, find candidates
4. Ask user: "Should I build this research capability using [actor name]?"
5. Build & register: Create script and add to registry

---

Scripts

Script	Purpose
run_actor.py	Run any Apify actor by ID
search_store.py	Search Apify store for actors
linkedin_jobs.py	LinkedIn Jobs search (uses vIGxjRrHqDTPuE6M4)

---

Integration with Outreach V2

For Tier 1 EDP data that requires external research:

/outreach-v2 → Phase 4 (List Criteria) → Identifies gap
                                        ↓
                           Apify skill fills the gap
                                        ↓
                           LinkedIn Jobs search → Company list
                                        ↓
                           Prospeo enrichment → Verified contacts

Example flow:

1. Tier 1 needs agencies posting "Operations Manager" roles
2. Apify LinkedIn Jobs actor searches for these postings
3. Returns company names with the EDP verified
4. Prospeo enriches with decision-maker contacts

---

Environment Variables

APIFY_API_KEY=your_api_key_here

Get your API key from: https://console.apify.com/account/integrations

---

Building New Research Capabilities

When you encounter a research need not in the registry:

Step 1: Search Store

uv run .claude/skills/apify/scripts/search_store.py --query "your research need"

Step 2: Evaluate Options

• Check pricing model (prefer pay-per-result)
• Check reviews and usage stats
• Verify output format matches needs

Step 3: Ask User

"I found [actor name] for this research need. It costs [pricing]. Should I add this to our research toolkit?"

Step 4: If Approved, Build Script

Create a purpose-built script in scripts/ and add to registry.

---

Quality Standards

Before using an actor:

• Verified pricing model (prefer pay-per-result)
• Tested with small sample
• Output format documented
• Added to registry if reusable

---

Related Skills

Skill	Relationship
find-leads	Apify provides EDP data, find-leads enriches contacts
pvp-list-criteria	Identifies when Apify research is needed
outreach-v2	Orchestrates the full pipeline