Advertising agent skills

A bunch of skills for spending paid media budget on purpose.

A collection of agent skills for paid advertising: channel selection, budget policy, audience design, creative development, and measurement, across search, paid social, video, native, and AI-answer surfaces.

Built for media buyers, performance marketers, PPC specialists, and growth leads running real budgets in B2B and B2C.

Related Collections

Other skills repositories I built for my colleagues at Nativa Labs:

• partnerships-skills: Partner ecosystem operations: for partner managers, BD leads, ecosystem heads
• revops-skills: Revenue operations: for RevOps managers, sales ops, marketing ops, CRM admins
• sales-skills: Sales execution: for SDRs, AEs, sales managers, heads of sales

Install

Install every skill in this repo, not just one. Skills here are atomic by design and reference each other freely: picking a single skill leaves its sibling skills uninstalled, so cross-references and routed handoffs go nowhere.

skills.sh (universal): works with any Agent Skills-compatible tool:

npx skills add mbfinotti/advertising-skills

Claude.ai:

1. add as a plugin marketplace: open Settings -> Capabilities -> Plugins
2. click Add -> Add marketplace -> Add from a repository
3. enter mbfinotti/advertising-skills
4. then Sync

Claude Code: install the plugin:

/plugin marketplace add mbfinotti/advertising-skills
/plugin install advertising-skills@advertising-skills

Codex (OpenAI): install via the Codex CLI:

codex plugin add github:mbfinotti/advertising-skills

Cursor: copy into Cursor's skills directory:

git clone https://github.com/mbfinotti/advertising-skills.git ~/.cursor/skills/advertising-skills

Cursor auto-discovers skills from .agents/skills/ and .cursor/skills/.

Gemini CLI: install as a Gemini extension:

gemini extensions install https://github.com/mbfinotti/advertising-skills

Update with gemini extensions update advertising-skills.

Skills

This collection covers the full advertising surface.

Start here

advertising-kickoff: Routes a paid advertising task to the right skill in this collection and bootstraps a shared project context artifact reused across sessions.

Meta

• advertising-career: Plans a paid media career from the candidate side: the junior-to-lead ladder, interview formats, an NDA-safe portfolio, and pay conversations.
• advertising-hiring: Builds the employer-side hiring packet for a paid advertising role: outcome scorecard, interview loop, work sample, and a 30-60-90 ramp.
• advertising-radar: Assembles a time-budgeted watch list of newsletters, podcasts, platform release notes, communities, and people, with a method for keeping it fresh.

Strategy & planning

• ad-platform-selection: Decides which paid channel families fit the business's economics, audience, funnel stage, and budget before any campaign gets built.
• ad-spend-allocation: Splits a fixed paid media budget across campaigns, platforms, funnel stages, and audiences based on expected marginal return.
• ad-format-fit: Flags ad formats mismatched to the objective, funnel stage, or production capacity, then separates a format problem from a creative one post-launch.
• paid-media-scaling: Decides when a proven campaign has earned a budget increase, how large each step should be, and what triggers a rollback.
• cac-roas-benchmark: Computes CAC and ROAS from real spend data and returns a healthy, watch, or unhealthy verdict against break-even, history, and external benchmarks.

Audience & targeting

• ad-audience-targeting: Turns an ICP and buying signals into layered audience tiers sized against platform floors, budgeted to the learning threshold, with exclusion rules.
• lookalike-audience-seeds: Selects and sizes the seed customer list behind a lookalike audience, with a consent gate and a fallback ladder when the match floor is missed.
• ad-negative-keywords: Mines search term reports into negative keyword lists, choosing match types, exclusion levels, and a review cadence that avoids overblocking converters.
• ad-buyer-group-mapper: Maps the buying committee for an offer and gives each role a messaging angle plus the targeting proxy that actually reaches them.
• retargeting-funnel: Designs a staged retargeting sequence with recency windows, behavioural depth tiers, a message and offer ladder, and per-stage frequency caps.

Creative

• ad-creative-brief: Turns a campaign goal and an audience insight into a brief a designer, editor, or creator can execute, with specs and a revision loop.
• ad-copy-variants: Produces genuinely distinct copy variants from one value proposition, each labelled by angle, awareness stage, and formula so results stay attributable.
• ad-hook-analyzer: Scores and force-ranks candidate video ad openings to decide which hooks deserve test budget, returning a shortlist rather than a performance prediction.
• ad-creative-test-plan: Designs a pre-launch creative test with a falsifiable hypothesis, per-cell budgets, required sample, and kill or scale rules registered in advance.
• ad-creative-fatigue: Decides whether a creative is genuinely wearing out or a confounder explains the decline, returning a verdict with confidence and the highest-return remedy.
• ugc-ad-scripts: Writes UGC-style short-form video scripts a creator can film - hook, body, CTA - with hook variants, delivery notes, and disclosure lines.
• conversational-ad-copy: Adapts messaging for one-shot ad slots inside AI assistant answers, covering recommendation register, situation targeting, grounding, and disclosure.
• thought-leadership-ads: Plans a campaign promoting an executive's existing organic posts as paid person-fronted ads: selection, exclusions, permission, budget, and measurement.
• ad-swipe-file: Builds a queryable library of competitors' running ads classified by format, hook, offer, and funnel stage, then converts it into ranked test hypotheses.

Campaign operations

• ad-bidding-strategy: Chooses the bidding policy per platform and goal, written up with an evaluation window, change discipline, and a rollback trigger.
• ad-budget-pacing: Tracks spend against budget and flags under-pacing or over-pacing early, reporting the pacing ratio, projected spend, and the corrective daily spend.
• ad-spend-guardrails: Sets top-level spend policy - allowable CAC, ROAS floor, kill-switch thresholds, owners, overrides - derived from margin, payback, and runway.
• ad-campaign-consolidation: Recommends which campaigns and ad sets to merge, which splits to preserve, and how to migrate without resetting learning across the account.

Measurement & diagnostics

• ad-conversion-tracking: Verifies conversion events fire once and deduplicate correctly before launch, then turns the evidence into a GO or NO-GO decision.
• ad-attribution-gap: Quantifies the discrepancy between platform, analytics, and CRM numbers, classifying every unit of the gap as timing, definitional, or unexplained residual.
• ad-account-diagnostic: Diagnoses the root cause of an underperforming ad account and returns a prioritised verdict with evidence and confidence, not a budget increase.
• paid-landing-page-audit: Audits a landing page taking paid traffic and returns a prioritised fix list covering message match, form friction, trust, speed, and mobile.

License

MIT © 2026 Maya-Beth Finotti