Author: Daniel · Primary KW: data collection tools (KD 4 · SV 1,000 · GSV 4,300 · CPC $2.00) Target persona: Data analytics managers / data engineers / ops teams whose budgets just got cut Funnel stage: Decision Draft v1 · skillsmp + ClawHub edition · 2026-04-28 Sources: 5 ClawHub BA skills + 5 skillsmp.com skills (verified via REST API 2026-04-28) --- 🚨 Your data team's budget got cut. Apify is $499/mo. Bright Data starts at $500. Octoparse Pro is $189. Zyte API is $450. Your stack used to be

🏆 1. ClawHub Web Data Skills (BA)

👉 https://clawhub.ai/

What it is: ClawHub is BrowserAct's skill marketplace — every skill on it runs on top of BA's stealth browser, which means: real browser fingerprints, automatic captcha handling, residential-IP rotation, all without you wiring infrastructure.

Why it's #1: When you collect data at scale, the bottleneck isn't "writing the scraper" — it's the stealth. SaaS tools charge $300+/mo for what's effectively a managed Chrome with rotation. ClawHub skills give you the same thing pay-per-call.

The 30-second recipe that works on most public sites:

browser-act stealth-extract \
  "https://target-site.com/page" \
  --fields "title,price,url,meta" \
  --output data.json

That's the universal entry point. The next 9 skills are when you need something more specific.

🥈 2. ClawHub Google Maps API Skill (BA)

👉 https://clawhub.ai/phheng/google-maps-api-skill

What it does: Pull POI data — name, address, rating, reviews, opening hours — for any geographic query. Without paying Google's $17/1K-call PSA pricing.

Why it matters: Most "data collection" projects start with location data. Restaurant lists, real estate comps, retail footprints — Google Maps is the fastest source if you don't pay enterprise rates.

Recipe:

browser-act stealth-extract \
  "https://www.google.com/maps/search/coffee+shops+austin" \
  --fields "name,address,rating,reviews,phone,hours"

3. data-collection-automation (skillsmp · wentorai · ★218)

👉 https://skillsmp.com/skills/wentorai-research-plugins-skills-research-automation-data-collection-automation-skill-md

What it does: Orchestrates multi-step data collection workflows — define a target, define the cadence, the skill schedules + retries + dedupes the runs.

Why it's on the list: 218 stars on skillsmp puts it in the top 0.01% of the marketplace. It's the closest open-source equivalent of Octoparse Cloud's "Scheduled Run" feature, except installed locally and called by your agent.

Install:

npx skills add wentorai/research-plugins/data-collection-automation

4. scrape (skillsmp · garrytan · ★98,391)

👉 https://skillsmp.com/skills/garrytan-gstack-scrape-skill-md

What it does: General-purpose web extraction. Point it at a URL, describe the fields, get JSON back.

Why it's on the list: 98,391 stars — the most-starred extraction skill on skillsmp by an order of magnitude. The "lingua franca" extraction skill that more specialized skills (like #3) often delegate to under the hood.

Install:

npx skills add garrytan/gstack/scrape

If you only install one extraction skill from the open-source side, install this one.

5. ClawHub Google News API Skill (BA)

👉 https://clawhub.ai/phheng/google-news-api-skill

What it does: Time-bounded news search → structured article list (title, source, snippet, date, URL).

Why it matters: News is the "freshness layer" of most data projects — competitive monitoring, brand tracking, regulatory updates. SerpAPI charges $50/mo for the same thing.

Recipe:

browser-act stealth-extract \
  "https://news.google.com/search?q=your+brand" \
  --fields "title,source,date,url,snippet"

6. data-collection-guide (skillsmp · orientpine · ★26)

👉 https://skillsmp.com/skills/orientpine-honeypot-plugins-isd-generator-skills-data-collection-guide-skill-md

What it does: Less an extractor, more a playbook. The skill walks the agent through choosing the right collection strategy: API vs. scrape vs. dataset vs. hybrid.

Why it's on the list: Most data collection projects fail at the design step, not the implementation step. Use this skill once before scope is locked — it'll save you from picking the wrong primary source.

Install:

npx skills add orientpine/honeypot-plugins/data-collection-guide

7. ClawHub YouTube Channel API Skill (BA)

👉 https://clawhub.ai/ccmagia2-gif/youtube-channel-api-skill

What it does: Channel metadata, subscriber count, video list, view counts — all without a YouTube Data API quota.

Why it matters: YouTube quota throttles at 10,000 units/day on the official API. One channel deep-dive can burn 200+ units. ClawHub's stealth path bypasses the quota entirely for moderate workloads.

Recipe:

browser-act stealth-extract \
  "https://www.youtube.com/@channel-name/about" \
  --fields "name,subscribers,videos,views,joined"

8. scrape-content (skillsmp · igor9silva · ★20)

👉 https://skillsmp.com/skills/igor9silva-meseeks-config-skills-scrape-content-skill-md

What it does: Article-content extraction. Hand it a URL, get the readable article back as clean Markdown — no nav, no ads, no boilerplate.

Why it's on the list: This is the skill version of what Mercury / Readability / Diffbot used to charge $200/mo for. Wire it after a SERP skill (like #5) to build a full "search → extract → summarize" pipeline.

Install:

npx skills add igor9silva/meseeks-config/scrape-content

9. learning-data-collection (skillsmp · majiayu000 · ★7)

👉 https://skillsmp.com/skills/majiayu000-claude-skill-registry-data-data-learning-data-collection-skill-md

What it does: ML training-data preparation. Splits raw collected data into train/val/test, normalizes schemas, generates the metadata file your downstream training script expects.

Why it's on the list: If your data collection feeds a model, you've got two jobs (collect + prep) that most teams treat as one and screw up. This skill enforces the boundary.

Install:

npx skills add majiayu000/claude-skill-registry/learning-data-collection

10. niche-data-collection (skillsmp · sellerai-com)

👉 https://skillsmp.com/skills/sellerai-com-sellerclaw-agent-agent-resources-agents-scout-skills-niche-data-collection-skill-md

What it does: Vertical-specific data collection scout. Given a niche keyword (e.g., "yoga mat for back pain"), the skill maps the relevant data sources, ranks them by quality, and produces a collection plan.

Why it's on the list: Useful as the "kickoff" skill on a fresh niche project — gives you a Plan B and Plan C if your first source goes dark.

Install:

npx skills add sellerai-com/sellerclaw-agent/niche-data-collection

⚠️ Reality check

You don't need:

❌ A $499/mo Apify subscription for managed Chrome rotation — ClawHub skills run on the same infra at pay-per-call rates
❌ A $189/mo Octoparse Pro seat for visual scraper builders your agent doesn't need
❌ A $450/mo Zyte API tier when 90% of your runs hit unauthenticated public pages
❌ 5 vendors solving 5 layers of one pipeline (search, extract, parse, dedupe, store)

You need:

✅ One stealth-extract skill (ClawHub root — skill #1 — for any new target)
✅ One playbook skill (orientpine data-collection-guide — skill #6 — for project kickoff)
✅ One general scraper (garrytan scrape — skill #4 — for the long tail)
✅ One specific data layer per vertical (Maps / News / YouTube depending on your work)
✅ A Claude or Codex agent to chain them

Monthly cost: ~$50 in pay-per-call usage.
Replaces: $1,000+/mo SaaS data stack.

Final thought

The data teams shipping insights in 2026 aren't the ones with the longest vendor list.

They're the ones who:

1. Picked 3 skills covering "extract / parse / orchestrate"
2. Wired them into one Claude agent
3. Spent the saved $11K/year on hiring an analyst — not paying for one more dashboard

Most teams won't do this. They'll keep paying Apify.

That's exactly why this works for the ones who do.

👉 Browse 5,000+ ClawHub data skills: https://clawhub.ai/
👉 Search 1.4M open-source skills on skillsmp: https://skillsmp.com/