agentleFS
adswipefileadswipefile/adswipefile-plugin/skills/weekly-competitor-review/SKILL.md

weekly-competitor-review

Skill0 starsChanged 27 days agoMITView on GitHub

Write a short weekly review of the new Meta ads the user's followed brands launched on AdSwipeFile, what changed against each brand's usual patterns, and three ideas to try, with a link to every ad. Runs by hand or as a scheduled Claude routine or ChatGPT task. Use for "weekly competitor review", "what did my competitors launch this week", "competitor ads digest" or a recurring competitor report.

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 2 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Keep the ads whose started date falls in the window.

    line 31
  2. Very specific:

    Read 3 to 6 of the newest filed ads that started in the window with get_ads, one call, spread across brands, for their hook, structure and …

    line 39
  3. Distinctive:

    When run as a scheduled routine or task: ask no questions, make no writes (no follows, saves or boards), and say which window the review …

    line 58
  4. Distinctive:

    No results: never spend, ROAS, conversions, or that an ad works or beat another.

    line 64
  5. Very specific:

    Keep calls few: one list_followed_brands, one list_followed_ads, at most 5 list_ads and one get_ads.

    line 67

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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

Weekly competitor review

Needs the AdSwipeFile connector. This skill reads the AdSwipeFile ad library of Meta ads through its MCP server. If no AdSwipeFile tools (search_library, list_ads) are available, say so before anything else and give the step for this host. Claude Code: claude mcp add --transport http adswipefile https://adswipefile.com/api/mcp. claude.ai or ChatGPT: add https://adswipefile.com/api/mcp as a custom connector. Other clients: https://adswipefile.com/docs/mcp?utm_source=agent-skill&utm_medium=referral&utm_campaign=weekly-competitor-review. Connecting opens a sign-in to an AdSwipeFile account. Steps that need only the user's own material can go on; never invent ads, brands or links in place of the library.

A short read of what followed brands launched in the window, set against what each brand usually runs. Words and evidence rules: the creative-strategy skill's REFERENCE.md.

The window is the last 7 days, or since the date the user gives.

Steps

  1. For an agency, one review per client: list_workspaces, then pass the client's workspace to list_followed_ads and list_followed_brands. get_usage says how many calls are left; when few are, write the review from step 2 alone and say so.
  2. list_followed_ads. Its following lists every follow, library brands and Facebook pages alike. If it is empty, the account follows nothing: explain that brands are followed on adswipefile.com/app/track (by name or Facebook page link), or offer follow_brand for brands the user names, and follow only with their go-ahead. Stop there.
  3. Keep the ads whose started date falls in the window. Each followed page also carries launched_last_7_days. Count per brand: ads, formats, and the hooks and offers their copy (title, body, cta) shows.
  4. list_followed_brands for the library slugs only (a page outside the library has none, so it has no filed ads to compare). For up to 5 library brands with new ads, in order of most new ads, list_ads with brand. The first page carries brand_patterns (most common structure, top visual format and hook tactic) to compare the new ads against, and lists the brand's filed ads newest first, with run.started.
  5. Read 3 to 6 of the newest filed ads that started in the window with get_ads, one call, spread across brands, for their hook, structure and offer. Skip this step when none are filed yet.

The review

Aim for under 350 words:

  • Headline: one line on the week.
  • Per brand: what launched (count, formats, hooks, offers) and what is new against its usual patterns, each point with a linked ad.
  • Across brands: patterns more than one brand shows.
  • Three ideas to try for the user's own ads, each tied to a linked ad.
  • Nothing new: the followed brands with no ads in the window, in one line.

End with the window covered and the line: this covers only followed brands' ads AdSwipeFile pulled; followed pages are checked weekly, so a brand's newest ad can be a few days old.

Run unattended

When run as a scheduled routine or task: ask no questions, make no writes (no follows, saves or boards), and say which window the review covered. When list_followed_ads shows no follows (neither brands nor pages), say so in one line and point to adswipefile.com/app/track.

Rules

  • No results: never spend, ROAS, conversions, or that an ad works or beat another. A long run is a reason to study an ad, never proof.
  • Link every ad by its url.
  • Keep calls few: one list_followed_brands, one list_followed_ads, at most 5 list_ads and one get_ads.

Done when the review is written with every cited ad linked and the window stated. Next: teach-ad-breakdown for an ad worth a closer look, or generate-ad-concepts to act on an idea.

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.