agricidaniel/claude-ads
20 skills16.2k installs116k starsGitHub
Install
npx skills add https://github.com/agricidaniel/claude-adsSkills in this repo
1Adsads is a Claude Ads companion skill that helps solo founders and small marketing teams reason about advertising platforms outside the core seven covered elsewhere in the agricidaniel/claude-ads family. It opens with a decision tree keyed to business versus consumer intent, visual discovery, brand awareness, and niche communities, then deep-dives Reddit and CTV/OTT with positioning, features, and rough benchmark ranges. The document is strategic reference material—use it when you already run or plan paid campaigns and need to justify testing Reddit conversation ads, premium CTV inventory, or directory placements like G2 without blindly copying Meta budgets. It assumes you will pair channel choice with creative, tracking, and landing-page readiness from your validate and ship work.1.1kinstalls2Ads Metaads-meta is an agent skill package for solo and indie builders who advertise on Meta’s family of apps and need more than ad-hoc chat advice. It performs a deep Meta Ads analysis aligned with the 2026 stack—Andromeda retrieval, GEM creative embeddings, and Lattice sequence optimization—so recommendations reflect how delivery actually ranks and filters creative today. The workflow evaluates roughly fifty checks spanning Pixel and Conversions API health, creative diversity and Entity-ID clustering risk, account and campaign structure, default Advantage+ Shopping Campaign and Advantage+ App Campaign patterns for sales, leads, and app goals, and audience targeting posture. It includes an Advantage+ assessment and creative-as-targeting scoring so you can see whether your assets match GEM’s embedding-driven targeting model. Invoke it when you are auditing an existing account, planning a relaunch, or debugging rising CPMs and weak learning before you increase budget. It is analysis and checklist driven—not a replacement for Meta’s UI or your media buyer’s judgment—and works best when you can share account facts, creative examples, or export context with your agent.986installs3Ads Googleads-google is a checker-style agent skill for solo founders and small teams running Google Ads who need audit discipline instead of ad-hoc tweaks. It prescribes a nine-step process: collect exports and change history, validate at least thirty days of data plus Search Terms Report coverage, load the 80-check google-audit reference, apply benchmarks and weighted scoring, mark every check PASS, WARNING, or FAIL, then emit a health score and action plan. Analysis spans conversion tracking (including Enhanced Conversions and Consent Mode v2), wasted spend, account structure, keywords, creative, and settings—with explicit coverage for Performance Max and AI Max features such as search-term matching, AI Brief, text customization, FUE, brand exclusions, and Smart Bidding signals. It is user-invokable false in metadata but still valuable as catalog guidance when you say Google PPC, PMax, or search terms audit. Supply account exports and reports; do not expect the skill to pull live Google Ads without your data.925installs4Ads Tiktokads-tiktok is an agent skill that deep-audits TikTok advertising for indie and solo operators running performance campaigns in Ads Manager, TikTok Shop, or Spark Ads workflows. You export account and pixel or Events API status, then the skill walks a documented 28-check rubric across creative quality, tracking, bidding, campaign structure, and shop integration, applying TikTok-specific benchmarks and platform specs before computing a weighted health score from zero to one hundred. It is built for builders who already spend on TikTok and need a structured pass before increasing budget, not for first-time channel strategy from scratch. The process expects you to read bundled reference audits and scoring rules rather than improvising checklist items in chat. Post–2026 USDS joint-venture nuances are included so compliance and shop setups are evaluated in current policy context. Use it when you say TikTok Ads, Smart+, creative diversity for retrieval, or TikTok Shop and want an action plan tied to failing checks.916installs5Ads Creativeads-creative is a marketing operations skill for indie builders running paid campaigns who need a disciplined creative audit instead of subjective thumb rules. The workflow collects assets or performance inputs per platform, loads platform-specs and benchmarks from the ads reference bundle, and applies the weighted scoring-system before scoring RSA headline depth, extension strategy, pin usage, and format diversity. It validates that fatigue calls cite real performance movement, compares consistency across networks, and ranks what to shoot or write next. Triggers include creative audit, ad fatigue, creative diversity score, and ad copy review. Use it during Grow when CTR softens or when launching new variants, and during Ship launch prep when you want compliance checks before spend goes live. It pairs naturally with competitor intelligence when you want external benchmarks alongside internal creative health.875installs6Ads Competitorads-competitor is an agent skill package for solo and indie builders who need structured competitor paid-media research without a dedicated ad-spy SaaS. It walks a repeatable workflow: name competitors, pull industry CPC, CTR, and conversion baselines from bundled references, then mine public ad libraries and transparency centers to inventory live creative, copy, formats, and geo signals. The skill emphasizes messaging themes, keyword posture, and rough spend indicators so you can see where rivals over- or under-invest. It closes with a competitive intelligence report that highlights whitespace you can test in validate or grow phases. Invoke when you say competitor ads, ad spy, Meta Ad Library, Google Ads Transparency, competitor PPC, or ad intelligence. It complements catalog creative and audit skills by focusing on external rivals rather than your own account health.874installs7Ads Landingads-landing is an agent skill for solo and indie builders running paid search or social campaigns who need a structured post-click audit instead of guessing why clicks do not convert. It walks through collecting URLs from active campaigns, loading benchmark and conversion-tracking references, then scoring each page on factors that matter for ads: message match between ad and above-the-fold experience, speed and mobile usability, trust cues, and form friction. The workflow emphasizes headline, offer, CTA, visual, and keyword consistency with the ad, because mismatch is treated as the top failure mode. Output is scored landing pages plus prioritized improvements so you can fix high-leverage CRO issues before increasing budget. It complements generic LP builders by focusing on campaign-specific quality, not brand storytelling alone.838installs8Ads PlanAds Plan is a marketing-agency paid advertising template packaged as agent procedural knowledge for solo builders and small agencies who manage more than one client. It encodes industry realities—different goals per account, white-label reporting, tight margins, and weekly platform changes—into a repeatable onboarding and setup path rather than a one-off campaign brief. The checklist walks discovery in week one (business type, goals, audience, competitors, budget, creative inventory, brand rules, and account access) and technical setup in week two (tracking audit, conversion events, server-side/CAPI where applicable, and analytics or CRM linkage). Use it when you are standing up or restructuring paid channels for a client or your own product and need a complete operational spine instead of scattered platform tips. It matters because botched tracking and weak onboarding in the first ninety days silently destroy measurable ROAS and client trust. Pair it with landing and legal pages before spend scales, and with analytics skills once campaigns are live.825installs9Ads AuditAds Audit is an advanced agent skill for solo founders and small growth teams running paid media across several networks at once. It orchestrates a full multi-platform advertising audit by collecting account evidence, validating that at least one platform has usable data, detecting business type, and delegating to six parallel platform-specific audit agents. Output includes a health score per platform plus an aggregate 0–100 score, framed by the repo’s 10-Principle Thinking Framework so analysis stays observational and prioritized rather than a shallow checklist. Trigger when users ask for a paid media audit, PPC audit, account health check, or full ad spend review. Amazon Ads, attribution, and server-side tracking are referenced as companion sub-skills for deeper parallel waves. The skill is marked not user-invokable directly in frontmatter—it behaves as an orchestrator best invoked from the broader claude-ads workflow in Claude Code.820installs10Ads Budgetads-budget is a paid-media agent skill for solo founders who run ads on multiple platforms and need a disciplined budget and bidding review instead of gut-feel tweaks. It collects performance and spend data, loads reference frameworks for allocation and bidding decision trees, benchmarks CPC and CPA, and applies a health-score style evaluation before recommending changes. Hard validation gates matter: it expects at least fourteen days of spend before judging kill or scale moves, and kill-list candidates need sufficient signal—roughly twenty clicks or one hundred dollars spent—so you do not pause winners on noise. Recommendations follow the 70/20/10 rule (proven, scaling, test), the 3x Kill Rule, and a 20% scaling rule for increments. Trigger it when you say budget allocation, bidding strategy, ad spend, ROAS targets, media budget, or scaling—even though it is not user-invokable by default in metadata, agents can run it inside a broader ads workflow. Best for indie SaaS, ecommerce, and content businesses already spending across channels.814installs11Ads Linkedinads-linkedin is a deep-analysis agent skill for solo B2B founders and marketers who run—or are considering—LinkedIn paid campaigns and need more than ad-hoc tips. It walks a structured process: collect Campaign Manager exports and Insight Tag status, run a full 27-check audit from linked references, compare results to LinkedIn-specific benchmarks, and emit a weighted health score with prioritized findings. Technical setup (Insight Tag, Conversions API, funnel events, CRM hooks) gets heavy weight because attribution failures silently waste indie budgets. The skill reflects LinkedIn’s late-2025 naming change so recommendations match what you see in the UI today. It is checker-oriented, not a campaign builder—invoke when you mention LinkedIn Ads, sponsored content, lead gen forms, InMail, ABM, or Thought Leader Ads and want an evidence-style report rather than creative brainstorming alone.806installs12Ads Youtubeads-youtube is a specialized analysis skill in the claude-ads package for solo builders and small teams running Google YouTube inventory. It ingests Google Ads data filtered to Video campaigns, then walks a fixed process: google-audit checks, platform video specs, benchmark tables, and a scoring-system health algorithm before reporting on setup, creative, targeting, and measurement. It explicitly addresses 2026-era changes such as Demand Gen replacing VAC, Shorts creative requirements, CTV shoppable formats, and frequency capping. Triggers include YouTube Ads, pre-roll, bumper ads, skippable in-stream, Shorts ads, Demand Gen, VAC, or CTV YouTube ads. The skill is not user-invokable by default in metadata, so agents typically reach it when the user names YouTube paid video work. Output is an evaluation-oriented repo/report style deliverable after campaign-type validation.804installs13Ads MicrosoftMicrosoft Ads Deep Analysis is a marketing agent skill for solo builders running Bing/Microsoft search, Performance Max, Audience Network, or Copilot-related placements. It follows a fixed process: collect account export and UET status, apply the 24-check microsoft-audit reference, score against benchmarks and the weighted scoring system, then output a health score and action plan. Emphasis sits on validating Google imports (a common indie mistake), syndication partner cleanup, bid targets often 20–35% below Google CPCs, and Microsoft-only audience features. Invoke when users mention Microsoft Ads, Bing PPC, UET, or import audits—not for generic Meta or Google-only playbooks without Microsoft context.771installs14Ads Generateads-generate is the image-production step in the claude-ads workflow for solo builders running paid campaigns without a design team. It reads structured inputs—campaign-brief.md and brand-profile.json—and calls banana-claude (via nanobanana-mcp) to emit ratio-correct creatives for platforms like Meta. Commands support full-batch generation, per-platform slices, and one-off prompt runs when you only need a single asset. The skill is explicit about environment setup: `/banana setup`, version floor v1.4.1+, and clear stop behavior when MCP image tooling is not configured. That makes it suitable for agent-driven launch sprints where copy and concepts already live in a brief from ads-create. After generation you can hand assets to upload workflows or further edit passes. Pair it with audit and brief skills so visuals stay on-brand and brief-compliant.757installs15Ads Createads-create packages campaign strategy into a machine-readable brief so solo builders can move from brand DNA to creatives without re-explaining positioning in every chat. It ingests brand-profile.json, optionally layers in audit findings, and outputs campaign-brief.md with concepts, messaging pillars, and platform-oriented copy. The skill stresses emotional pull and spec compliance through its 10-Principle Thinking Framework—headlines and CTAs should sound like the brand, not generic filler. CLI-style invocations support narrowing platforms or optimizing for objectives such as lead generation. This is the planning artifact that ads-generate expects, which keeps copy and visual prompts aligned. Use it when you are ready to commit to paid distribution but still need structured concepts before spending on design or API image generation.756installs16Ads DnaAds DNA is an agent skill that turns a company website into a structured brand profile for paid advertising. Solo and indie builders run it when they need brand-consistent ad creative without manually copying style guides or guessing colors from screenshots. The skill prompts for a URL if missing, fetches pages with WebFetch, observes visual and verbal patterns on the site, and emits brand-profile.json that Claude Ads creation and generation skills consume. It aligns extraction with listen-first principles from the ads thinking-framework reference so classifications match what the site actually communicates. Use it whenever triggers mention brand DNA, brand voice, style guides, or analyzing brand identity before campaigns—not as a general SEO or logo design tool.753installs17Ads PhotoshootAds Photoshoot is an agent skill for solo founders and marketers who need professional product imagery without a studio budget. Starting from a local image, URL, or text description, it runs a structured photoshoot workflow that produces five distinct visual treatments—Studio, Floating, Ingredient, In Use, and Lifestyle—each rendered at ad-friendly dimensions. The quick-reference table documents slash commands for interactive runs, style filtering, and multi-platform size generation. Environment setup is explicit: banana-claude at v1.4.1 or newer with nanobanana-mcp configured through `/banana setup`. The process begins by collecting product inputs in a single conversational turn, then drives generation toward channels that expect square and vertical assets. It is not a general-purpose image editor; it is optimized for performance creative pipelines where consistent product presentation across styles speeds up test-and-scale ad workflows in Claude Code-style environments.729installs18Ads Appleads-apple is an agent skill for solo mobile app founders and small growth teams who advertise on Apple Ads (rebranded from Apple Search Ads in April 2025). You feed dashboard exports or pasted metrics; the skill walks a defined process: detect active placements, score campaign structure and bid health, evaluate Custom Product Pages, AdAttributionKit view-through windows, MMP attribution, budget pacing, Today/Search/Product Pages TAP coverage, and Maximize Conversions settings against country-level goal CPA benchmarks. Output is a findings report with PASS, WARNING, and FAIL labels plus an action plan—useful when you are launching an iOS app or growing paid installs without hiring a dedicated ASA agency. Trigger phrases include Apple Ads, ASA, App Store ads, and AdAttributionKit. It is analysis and checklist discipline, not automatic campaign creation inside Apple’s console.725installs19Ads Mathads-math is an agent skill that walks solo and indie builders through pay-per-click financial modeling using only pasted spend, conversion, and revenue data—no Google Ads API required. It detects or asks which calculator you need, collects inputs for the same time period, runs CPA, ROAS, CPL, break-even, budget forecast, impression-share sizing, LTV:CAC, and MER, and returns interpreted results with actionable recommendations. It fits builders who export campaign stats manually and want spreadsheet-grade rigor inside Claude Code, Cursor, or Codex without standing up integrations. Use it during pricing validation, before scaling launch budgets, and while reviewing grow-phase performance to see if paid channels clear margin targets. The workflow emphasizes transparent formulas and benchmark callouts so you can paste a row from Ads Manager or a CSV fragment and get a decision-ready read in minutes.572installs20Ads Testads-test is a Claude skill for A/B test design and experiment planning on paid advertising channels. Solo builders running their own Meta, Google, or LinkedIn spend use it when they mention split tests, hypotheses, statistical significance, sample size, or how long to run an experiment. The workflow starts by clarifying what changes—creative, audience, bidding, or landing page—then forces a structured IF-THEN-BECAUSE hypothesis, calculates required sample size and estimated duration, and points to platform-specific experiment UIs. It includes a hypothesis quality checklist so you isolate one variable instead of confounding results. For indie SaaS or ecommerce founders without a growth team, the skill turns ad hoc button-clicking into a measurement plan with defined success criteria before budget burns. It complements creative and targeting skills by making incrementality and readouts explicit.568installs