You give AdWhispr a brand name or a keyword and it searches the Meta Ad Library, then hands the results back as structured data your agent can actually work with: each ad classified by hook, format, and offer, and the set ranked by how long each has been running. It is the raw research primitive the deeper skills build on. It never invents spend or a click rate; the ranking is longevity, because that is the one honest signal the ad library exposes.
Use this when
You need to look something up fast, a brand, a product term, a category phrase, and want clean, ranked results rather than the raw ad library interface. It is the right entry point when you are exploring, checking whether a brand advertises at all, or feeding results into a larger analysis.
Do not use this when you want a full read on one brand's strategy; the Competitor Ad Teardown skill goes deeper. And do not use it to compare two brands; the Head-to-Head Brand Comparison skill is built for that.
What you need
- A brand or page name, or a keyword or phrase to search across advertisers.
- AdWhispr connected to your AI client. The one-time setup is in the box at the top of this page. No ad account of your own is needed for read-only research.
- The market, if it matters, since the ad library is region aware and results differ by country.
How it runs
- Confirm the query with the user: a brand name, or a keyword and the niche it belongs to, plus the market if it is region sensitive.
- Search the Meta Ad Library for matches, resolving brand names to advertisers where the query is a brand.
- Pull the matching ads and read each one's first-seen and last-seen dates.
- Rank the results by days running, longest first, so the durable ads surface ahead of the fresh ones.
- Return structured results, each ad classified by hook, format, and offer, with days running attached, ready to feed into a deeper skill.
What the skill checks
- Results are read fresh from the tracked library, not carried from an earlier session, since advertisers pause and launch between searches.
- No spend, CTR, or impression figure is attached to any result. Days running is the only performance quantity, and none of the others are public.
- The market filter is confirmed before returning an empty result, since no matches under the wrong region is a false negative, not a finding.
- A brand-name query is resolved to the correct advertiser before results are returned, so a similarly named page is not mistaken for the target.
When it stops
- Done: structured, longevity-ranked results for the query, each ad classified and dated.
- Nothing to report: the search returns no ads in the requested market, which is reported plainly with the date.
- Blocked: the brand name does not resolve to any advertiser after reasonable variants are tried, so the skill asks you for the exact name or URL.
What it will and won't do
Read only. This skill searches the public ad library and returns structured results. It never touches an ad account, never spends, and never contacts anyone. Acting on a result, cloning an ad or launching against it, is a separate skill with its own approval.
What you get
Clean, structured search results from the Meta Ad Library: the matching ads ranked by days running, each classified by hook, format, and offer, ready to read directly or hand to a deeper analysis skill. No spend estimates, no invented scores, no ranking by anything but how long each ad has stayed live.
FAQ
How do I search the Facebook Ad Library from my agent?
Give the skill a brand name or a keyword and it searches the Meta Ad Library, then returns structured, ranked results your agent can use directly. It resolves brand names to advertisers and ranks every result by how long the ad has been running.
Can I search by keyword instead of by brand?
Yes. Search a product term or category phrase and the skill returns matching ads across advertisers, ranked by longevity, so you see the durable creative in the space rather than one brand's library.
Why are results ranked by how long an ad has run?
Because that is the only honest performance signal the ad library exposes. Spend and click rates are not public, so the skill ranks by days running, the closest proxy for what is working, rather than inventing numbers.
Does this touch my ad account?
No. It searches public ad-library data and returns results. Your own account is never queried by this skill.