agentleFS
adswipefileadswipefile/adswipefile-plugin/skills/creative-strategy/SKILL.md

creative-strategy

Skill0 starsChanged 27 days agoMITView on GitHub

Start here for creative strategy on Meta ads with AdSwipeFile as the evidence. Routes research, ideation, making, review and learning to the right skill, from mapping competitors to concepts, scripts and audits. Use for "help me with creative strategy", "where do I start with my ads", "build a creative strategy for my brand", ad concepts, competitor research or a creative brief.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    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:

    Every week, weekly-competitor-review, by hand or as a scheduled task.

    line 54
  2. Very specific:

    Every cited ad is linked by its url.

    line 64
  3. Very specific:

    Before a long run of calls (a batch of concepts, several competitors), get_usage says how many are left, at no cost.

    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

Creative strategy

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=creative-strategy. 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.

These skills work one loop: research, ideate, make, review, learn. Two kinds of material feed it:

  • Other brands' public Meta ads, from the AdSwipeFile tools: copy, tags, breakdowns, run dates. They show how ads are built, never how they performed.
  • The user's own material: product facts, customer reviews, their ads and their results. It comes from the user: pasted, uploaded, linked, or from another connector they have. Ask for it plainly; never invent it.

Words, evidence rules and the tools by job live in REFERENCE.md. Read it once before the first skill.

Skills

StageSkillEnds in
Researchmap-competitorsA competitor map: direct, indirect, admired
Researchtrack-competitor-adsCompetitors followed on AdSwipeFile, a first read per brand
Researchweekly-competitor-reviewA short weekly review of what followed brands launched
Researchcompetitor-adsOne brand's formats, hooks, angles and offers, linked
Researchmap-personasPersonas with anchors, trigger moments and the user's customer words
Researchfind-winning-adsLinked reference ads for a niche, format or keyword
Ideategenerate-ad-conceptsDraft concepts, one per anchor, persona and moment
Ideatepick-creative-mechanicThe mechanic each concept uses, with reference ads
Ideatewrite-hooks-from-adsOriginal hooks and headlines in patterns from real ads
Reviewaudit-ad-creativeFeedback on one ad and an iteration brief
Learnteach-ad-breakdownA breakdown of one ad and the takeaway to reuse
Learnreview-saved-adsWhat the user's saved ads have in common
Learnrecord-creative-learningA scoped learning from the user's own results

Usual order for a new brand

  1. map-competitors, then track-competitor-ads to follow them and read what they keep running.
  2. map-personas from the user's reviews and customer words.
  3. generate-ad-concepts, picking a mechanic per concept with pick-creative-mechanic and hooks with write-hooks-from-ads.
  4. audit-ad-creative on what is about to ship.
  5. Once the user has results from their own ads, record-creative-learning.
  6. Every week, weekly-competitor-review, by hand or as a scheduled task.

Pick up at any stage the user is already at. A single question ("what is this brand running?") goes straight to its skill.

Rules

  • Outputs go in the reply, or in a file when the client can write files. Keep references on AdSwipeFile (save_ad, add_ad_to_board, create_board, set_ad_note) only when the user asks or agrees.
  • Every cited ad is linked by its url. The library holds no spend, ROAS or conversions: a long run is a reason to study an ad, never proof it worked.
  • Results-type calls count against the user's calls: read samples with get_ads, keep calls few. Before a long run of calls (a batch of concepts, several competitors), get_usage says how many are left, at no cost.
  • A list or search cut short answers with a note: pass it on once with its link, work from what came, and narrow with filters instead of paging.
  • One client or brand among several (an agency): list_workspaces once, then pass that client's workspace to every save, follow and board, so each client's research stays apart.
  • A board of references is the deliverable a client or team sees: once it is filed, offer share_board for a read-only link.
  • A brand or result the library is missing: offer to report it with submit_feedback, once, with the user's go-ahead.

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.