agentsclimarketplace
Sign in

Assets

Skill chanktb/claude-google-ads/skills/assets

Generates Google Ads creative assets shared across campaign types: responsive search ad headlines and descriptions, Performance Max asset-group copy and search themes, Demand Gen image/video briefs, and sitelinks/callouts. Enforces character limits, brand voice from account-context.yaml, and policy/ trademark safety. Use when the user says "ad copy", "headlines", "descriptions", "assets", "creative", "rsa", "asset group copy", "sitelinks", "callouts".From its SKILL.md

Install
npx -y skills add chanktb/claude-google-ads --skill assets

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 10 stars10 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

3.7 KB, 802 tokens by cl100k_base, as published. Nobody here has run it

Google Ads — Assets (shared creative)

One creative generator used by every builder, so copy rules and voice live in one place. Reads account-context.yaml (voice, brand, offers, vertical) and the data source (best sellers, real prices).

Operating rules

  • Pull voice/positioning/offers from context — no generic filler, no off-brand claims.
  • Enforce character limits and policy/trademark rules. Never put a competitor trademark in ad text.
  • Use REAL product names/prices/stock from the data source; don't invent.

Model dispatch (run cheap, decide expensive) — see ${CLAUDE_PLUGIN_ROOT}/references/model-tier-dispatch.md

  • Scout (haiku) — pulling real product names/prices/stock from the data source; the Google-accurate char-count + ready-to-copy wrapping pass.
  • Judge (main session) — all copywriting (headlines, descriptions, long headlines, search themes, sitelinks, briefs), voice/positioning, policy/trademark safety, pin discipline. Creative is the one thing that never goes to a cheap tier; only the inputs (product facts) and the validation (char count) do.

STEP 1 — RSA (Search / Branded)

Per ad group: up to 15 headlines (≤30) and 4 descriptions (≤90), themed to the ad group's intent. Mix angles: keyword/intent · offer (shipping, selection, in-stock) · qualifier (professional/wholesale/ bulk or value, per vertical) · best-seller names · trust/social proof · CTA. Pin only must-include/legal claims (over-pinning kills RSA flexibility). Aim for Good/Excellent ad strength.

STEP 2 — PMax asset-group copy

Per asset group: 15 headlines (≤30), 5 long headlines (≤90), 5 descriptions (≤90; #1 ≤60), and up to 50 search themes (raised from 25 in 2025; distinct per AG, no overlap). Use the angle mix in ${CLAUDE_PLUGIN_ROOT}/references/pmax-best-practices.md.

STEP 3 — Demand Gen briefs

Image + video briefs by spec (sizes + shot list + CTA). Combined image+video outperforms video-only. Provide concrete direction (subject, context, composition, lighting, on-screen text/CTA) per format.

STEP 4 — Extensions

Sitelinks (text ≤25; two descriptions ≤35 each; live URL), callouts (≤25), structured snippets (value ≤25). Level depends on campaign type: in PMax these attach at the campaign level (spec.extensions), never the asset group. In Search, they may be campaign- or ad-group-level (ad-group only when groups differ enough to warrant distinct links). See builder-pmax STEP 7b.

Validation

Count characters the Google way (NFC-normalized, CJK/full-width = 2, no trailing/hidden whitespace) per ${CLAUDE_PLUGIN_ROOT}/references/google-ads-formatting.md — not naive len(), which lets copy pass here but get rejected on entry. Flag off-brand/banned phrasing or competitor trademark in text; confirm product references are real (in stock, correct price). Keywords/negatives come back ready-to-copy wrapped ([exact] / "phrase" / broad). Return assets keyed by campaign type + group for the builder to drop into campaign-spec.json.

To build / refine later

  • Optional AI image generation hook for Demand Gen / PMax visuals.
  • Shared char-limit/policy validator reused by builders.

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Gives 0 of the 12 instructions most marketing audience skills give in 802 tokens

Counted across 690 of the 894 authors here whose files we hold, read 2026-08-07

  • Apply Poppins font to headingsin 41 of 690, across 6 files
  • Apply Lora font to body textin 41 of 690, across 6 files
  • Use Arial fallback for headingsin 39 of 690, across 4 files
  • Use Georgia fallback for body textin 39 of 690, across 4 files
  • Maintain text hierarchy and formattingin 39 of 690, across 4 files
  • Use accent colors for non-text shapesin 38 of 690, across 3 files
  • Use RGB values for precise color matchingin 38 of 690, across 3 files
  • Use brand colors for primary text and backgroundsin 36 of 690, across 1 file
  • Read product marketing context file before asking questions, starting, or auditingin 35 of 690, across 23 files
  • Use active voice instead of passive voicein 26 of 690, across 10 files
  • Implement or generate appropriate JSON-LD structured datain 24 of 690, across 17 files
  • Prioritize clarity over clevernessin 22 of 690, across 8 files

Said here and by no other author read

  • pull voice positioning offers from context
  • use real product names prices stock
  • enforce character limits and trademark rules
  • write all copy in the main session
  • dispatch product fact extraction to cheap tier
  • dispatch character count validation to cheap tier

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.