agentleFS
Sign inStart free
HumanleapHumanleap/agent-skills/skills/ad-library-scraper/SKILL.md

ad-library-scraper

Skill0 starsChanged 41 days agoMITView on GitHub

Search public ad libraries through ToolRouter's MCP server - Meta (Facebook and Instagram) Ad Library, Google Ads Transparency Center, LinkedIn, TikTok and Reddit ads. Use when the user wants to see a competitor's running ads, ad copy and creative, how long ads have run, which platforms they run on, or video ad transcripts.

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 3 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read the public ad libraries that Meta, Google, LinkedIn, TikTok and Reddit publish, from one MCP connection.

    line 9
  2. Distinctive:

    No guessing at spend: Only report spend or reach when the library returns it.

    line 20
  3. Very specific:

    Call these through use_tool with tool, skill and input.

    line 51
  4. Distinctive:

    Channel mix check: run the same brand through Meta, Google, LinkedIn and TikTok to see where it actually advertises.

    line 57

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Ad Library Scraper

Read the public ad libraries that Meta, Google, LinkedIn, TikTok and Reddit publish, from one MCP connection. Returns ad copy, creative URLs, start dates, platforms and, where the library publishes it, reach and spend.

Installation

Install with pnpm dlx skills add Humanleap/agent-skills --skill ad-library-scraper. Connect the client's remote MCP server to https://api.toolrouter.com/mcp (Claude: Customize → Connectors → Add custom connector; Claude Code: claude mcp add toolrouter -- npx -y toolrouter-mcp). Setup: https://toolrouter.com/docs/markdown/quickstart.

Hard Rules

  1. Resolve the advertiser first. Keyword search matches any ad that contains the word. To see one company's ads, look up its page or advertiser ID, then fetch that company's ads.
  2. Use the exact operation. Pick the tool and skill below, or call discover. Never invent an input field.
  3. State the window. Say which country, status and date range the results cover.
  4. No guessing at spend. Only report spend or reach when the library returns it.

Authentication and Cost

The server provisions an anonymous account on first use. Paid calls need credits: if use_tool returns a claim link, give it to the user so they can sign in and add credits. Most ad library calls cost $0.005; discover shows the live price. get_google_company_ads with get_ad_details: true costs more per ad, so leave it off until the user needs full detail.

Core Workflow

  1. Find the advertiser. Meta: search_facebook_ad_companies → page_id. Google: search_google_advertisers → advertiser_id, or pass the domain.
  2. Pull their ads. get_facebook_company_ads with page_id, country, status:"ACTIVE" and trim:true. get_google_company_ads with domain or advertiser_id.
  3. Open the winners. get_facebook_ad with get_transcript:true for a video ad's script; get_google_ad for one Google creative.
  4. Summarise. Table of start date, days running, platforms, format, hook (first line) and call to action. Long-running ads are the ones the advertiser keeps paying for.

Essential Operations

All tool: "ad-library-search".

JobskillInput
Find a brand's Meta page IDsearch_facebook_ad_companies{"query":"Notion"}
A brand's active Meta ads in one countryget_facebook_company_ads{"page_id":"825120160874613","country":"GB","status":"ACTIVE","trim":true}
Meta ads mentioning a keywordsearch_facebook_ads{"query":"protein bar","country":"US","search_type":"keyword_exact_phrase","trim":true}
One Meta ad, with video transcriptget_facebook_ad{"url":"https://www.facebook.com/ads/library/?id=123","get_transcript":true}
Find a Google advertisersearch_google_advertisers{"query":"Shopify"}
A domain's Google adsget_google_company_ads{"domain":"shopify.com","region":"US"}
One Google ad (text via OCR)get_google_ad{"url":"https://adstransparency.google.com/advertiser/AR.../creative/CR..."}
LinkedIn ads by companysearch_linkedin_ads{"company":"Salesforce"}
TikTok ads by advertisersearch_tiktok_ads{"advertiser_name":"Duolingo"}
One TikTok ad (hook, landing page, video)get_tiktok_ad{"ad_id":"<id or library.tiktok.com URL>"}
Reddit promoted postssearch_reddit_ads{"query":"project management"}

Call these through use_tool with tool, skill and input.

Common Patterns

  • Competitor ad teardown: resolve 3 competitors' Meta page IDs, pull active ads in the user's market, sort by start date, and compare hooks, offers and formats.
  • Creative brief: take the 5 longest-running ads, pull transcripts for the video ones, and list the hooks, proof points and calls to action that repeat.
  • Channel mix check: run the same brand through Meta, Google, LinkedIn and TikTok to see where it actually advertises.
  • Landing page follow-through: pass an ad's landing URL to web-scraper/scrape_page to read the offer behind the ad.
  • Organic vs paid: pair with the social-media-scraper skill to compare what a brand promotes with what it posts.

Gotchas

  • search_facebook_ads is a keyword search across all advertisers: "notion" also matches unrelated ads that use the word. Resolve the page ID for brand research.
  • Several pages can share a brand name. Pick the one with the matching Instagram handle, follower count or page alias.
  • Meta results default to active ads; set status:"ALL" for history. Use trim:true to keep responses small, and cursor to page.
  • An empty Google result for a domain can mean the advertiser is registered under another name. Try search_google_advertisers.
  • Pass query or advertiser_name to search_tiktok_ads, not both.

Quick Reference

Connect → resolve page/advertiser ID → company ads (country, active, trim) → open top ads → table of hooks, dates and platforms.

More: https://toolrouter.com/tools/ad-library-search

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free