/competitor-signals

name: competitor-signals
description: Extract leads from competitor product activity — Product Hunt commenters/upvoters, HN posts about competitors, case studies, testimonials, tech press, and switching signals. Detects people actively switching from competitors as highest-priority leads.
user-invocable: true
allowed-tools: Bash, Read, Write, Edit, Grep, Glob, WebFetch, WebSearch
argument-hint: [config-json-path]

Competitor Signals

Find leads by monitoring competitor product activity. Instead of looking for your prospects directly, watch your competitors' audience — every person engaging with a competitor launch is self-identifying as in-market for your category.

When to Use

• User wants to find people engaging with competitor products
• User mentions Product Hunt launches, competitor press coverage, or competitor case studies
• User wants to find people switching from or evaluating competitor products
• User asks "who is using [competitor]" or "who is looking at alternatives to [competitor]"
• User wants to monitor competitor activity for lead generation
• User has a clear list of competitors and wants to mine their audience

Prerequisites

• Python 3.9+ with `requests` and optionally `python-dotenv`
• Product Hunt developer token (free, optional — get at `api.producthunt.com/v2/oauth/applications`)
• Apify API token in `.env` (fallback for PH if API names are redacted, optional)
• Working directory: the project root containing this skill

Phase 1: Collect Context

Step 1: Gather Competitor Information

Ask the user:

> "To find leads from competitor activity, I need: > 1. **Who are your competitors?** (product names and company names) > 2. **Do you know their Product Hunt slugs?** (the URL path on producthunt.com/posts/SLUG) > 3. **Any specific competitor launches or announcements you've seen recently?** > 4. **Are there competitors or signals you specifically want to track?** (e.g., a competitor just raised funding, launched a new feature, or got press coverage)"

Step 2: Discover Competitors (if user needs help)

If the user doesn't have a complete competitor list, help them discover competitors:

**2a. Product Hunt search:**

• Search producthunt.com for the user's product category
• Note: PH doesn't have a great search API — use web search: "site:producthunt.com [product category]"

**2b. G2/Capterra category pages:**

• Search: "[product category] G2" or "[product category] Capterra"
• These pages list all competitors in a category with rankings

**2c. "Alternatives to" sites:**

• Search: "[known competitor] alternatives"
• Sites like alternativeto.net, slant.co, stackshare.io list competitors

**2d. Ask the user:** > "Based on my research, here are competitors I've found in your space: [list]. Are there any I'm missing? Any you'd like to exclude (e.g., not really competitors, too different in market segment)?"

Step 3: Find Product Hunt Slugs

For each competitor, find their PH launches:

• Search: "site:producthunt.com [competitor name]"
• Or browse: `producthunt.com/products/[competitor-name]`
• Note the slug from the URL: `producthunt.com/posts/SLUG`
• A competitor may have multiple launches (initial launch + feature launches)

Step 4: Identify Competitor Web Pages to Scrape

For each competitor, identify pages the agent should scrape:

**Case studies page:** `[competitor].com/customers` or `[competitor].com/case-studies`

• Extract: company names, logos, quotes, person names, titles
• These are PROVEN BUYERS in the category

**Testimonials page:** Often on the homepage or a dedicated page

• Extract: person name, title, company, quote
• These are current users who publicly endorsed the competitor

**Blog:** `[competitor].com/blog`

• Guest posts by customers are case studies in disguise
• "How [Company X] uses [Competitor]" = case study

Present all discovered pages to the user for review.

Phase 2: Agent-Driven Scraping

Step 5: Scrape Competitor Websites

Before running the tool, the agent should manually scrape competitor case studies and testimonials. This is agent-driven because every competitor website has a different format.

**For each competitor's case study page:** 1. Navigate to the page using web fetch or Chrome DevTools 2. Extract all customer company names and any associated person names/quotes 3. Note the case study URL for reference

**For each competitor's testimonials page:** 1. Extract: person name, title, company, quote text 2. These are high-value signals — these people actively chose to endorse the competitor

**Save all scraped data** to `${CLAUDE_SKILL_DIR}/../.tmp/competitor_manual_signals.json`:

[
    {
        "person_name": "Sarah Chen",
        "company": "TechCorp",
        "signal_type": "case_study_company",
        "signal_label": "Competitor Case Study",
        "competitor": "Twilio",
        "context": "How TechCorp scaled video calls to 100K users with Twilio",
        "url": "https://twilio.com/case-studies/techcorp",
        "profile_url": "",
        "date": "",
        "source": "Manual",
        "engagement": 0
    }
]

Step 6: Check Tech Press

Search for recent articles about competitors:

• "[competitor] TechCrunch"
• "[competitor] The New Stack"
• "[competitor] InfoQ"
• "[competitor] DevOps.com"
• "[competitor] launch announcement"
• "[competitor] raises funding"

For articles found:

• Note the article URL and key companies/people mentioned
• If the article has comments, check for people expressing opinions
• Add notable findings to the manual signals JSON

Phase 3: Execute Tool

Step 7: Save Config

cat > ${CLAUDE_SKILL_DIR}/../.tmp/competitor_signals_config.json << 'CONFIGEOF'
{
    "competitors": ["Twilio", "Agora", "Vonage", "Daily.co"],
    "product_hunt_slugs": ["twilio-video", "agora-2", "daily-co"],
    "days": 90,
    "manual_signals_file": "${CLAUDE_SKILL_DIR}/../.tmp/competitor_manual_signals.json",
    "skip": []
}
CONFIGEOF

Step 8: Run the