Agent skill
meta-ads-builder
Build Meta/Instagram ad campaigns with audience targeting, ad copy, and creative specs
Filed under ABM and paid.
From ekatasingh1107/b2b-gtm-skills · 99 skills · 2 · pushed 2026-04-11
What it does when it runs
Build Meta/Instagram ad campaigns with audience targeting, ad copy, and creative specs
Read from the skill and the 1 file bundled beside it. A skill’s own description is written to be selected by an agent, so it describes the job and not the dependencies.
- Keys and connectors you must supply
- None found.
- Hosts it reaches
- No third-party host appears in the skill or its bundled files.
- Tool permissions it declares
- No
allowed-toolsin the frontmatter. It does act, so it runs under whatever permissions your session already grants. - Actions present in the files
- network
Install it
View source on GitHub ↗git clone --depth 1 --filter=blob:none --sparse https://github.com/ekatasingh1107/b2b-gtm-skills.git /tmp/b2b-gtm-skills git -C /tmp/b2b-gtm-skills sparse-checkout set "skills/composites/meta-ads-builder" mkdir -p ~/.claude/skills/meta-ads-builder cp -R "/tmp/b2b-gtm-skills/skills/composites/meta-ads-builder/." ~/.claude/skills/meta-ads-builder/
Picked up without a restart. A project skill of the same name is shadowed by your personal one. For one repository only, swap ~/.claude/skills for .claude/skills. Claude Code docs ↗
The folder is the same in every client that implements the format — 46 of them — so if yours is not above, only the destination changes.
The skill
Source on GitHub ↗Reproduced in full from ekatasingh1107/b2b-gtm-skills/blob/eae8dd0bb98da1c8e84abd297066a87015dd860f/skills/composites/meta-ads-builder/SKILL.md, which is licensed MIT (repository). 1,665 words, 27 headings.
Meta Ads Builder
Builds Meta (Facebook/Instagram) ad campaigns end-to-end: defines campaign objectives, builds audience targeting (interests, behaviors, lookalikes, custom audiences derived from ICP), generates ad copy variants for each placement, recommends creative formats and specs, defines budget allocation, and outputs a complete campaign blueprint ready for setup in Meta Ads Manager.
Prerequisites
- WebSearch tool available for audience research and competitor ad analysis
- Business services, target audience, and budget range
- Optional:
agency.config.jsonfor ICP, services, and brand voice - Optional: Meta Ad Library access via WebSearch for competitive intelligence
Capabilities Used
ad-creative-intel-- for winning creative angles and hook patternsad-spy-- for competitor ad intelligencelanding-page-auditor-- for ad-to-landing-page consistency
Phase 0: Read Config
- Read
agency.config.jsonfrom the project root (if available). - Extract
icp.segments[]for audience targeting: job titles, industries, company sizes, pain points. - Extract
services[]for offer messaging. - Extract
case_studies[]for social proof in ad copy. - Extract
brand_voicefor tone and style guidelines. - Check
tools.websearchavailability. - Accept parameters:
business_name-- (required) business nameobjective-- (optional)awareness|consideration|conversion|lead_gen. Default:lead_genservices-- (optional) array of services to promote. Default: derive from configlanding_page-- (optional) destination URLbudget_monthly-- (optional) monthly budget in local currencytarget_market-- (optional) geographic targetingexisting_assets-- (optional) list of available creative assets (images, videos)offer-- (optional) specific offer or lead magnet to promote
Phase 1: Campaign Objective Selection
Objective Mapping
Based on the goal, recommend the appropriate Meta campaign objective:
| Business Goal | Meta Objective | Optimization Event | Best For |
|---|---|---|---|
| Brand awareness | Awareness | Reach / Ad recall lift | New market entry, top of funnel |
| Website traffic | Traffic | Link clicks / Landing page views | Content promotion, blog traffic |
| Lead generation | Leads | Lead form / Conversions | B2B leads, service inquiries |
| Conversions | Sales | Purchase / Add to cart | Ecommerce, direct response |
| Engagement | Engagement | Post engagement / Video views | Community building, social proof |
Campaign Structure
Campaign: [Objective] - [Service/Offer]
Ad Set 1: Cold - Interest Targeting
Ad Set 2: Cold - Lookalike Audience
Ad Set 3: Warm - Website Retargeting
Ad Set 4: Hot - Engaged Retargeting
Phase 2: Audience Targeting
Cold Audiences (Prospecting)
Interest-Based Targeting Build interest targeting from ICP:
- Job title interests (CEO, Founder, Marketing Director, Ecommerce Manager)
- Industry interests (ecommerce, D2C, retail, specific verticals)
- Tool/platform interests (Shopify, WooCommerce, Magento -- indicates ecommerce involvement)
- Publication/thought leader interests (industry blogs, influencers)
- Behavior signals (business page admins, small business owners, online purchasers)
- Narrow with AND conditions to improve quality
Lookalike Audiences Recommend lookalike sources:
- Website visitors (past 90 days) -- 1% lookalike
- Email list / CRM contacts -- 1% lookalike
- Lead form completers -- 1% lookalike
- Top customers (by LTV) -- 1% lookalike
- Page engagers -- 2-3% lookalike (wider)
- Video viewers (50%+) -- 2% lookalike
Demographic Filters
- Age range: based on ICP (typically 25-55 for B2B)
- Location: target market geography
- Language: as appropriate
- Exclude existing customers and recent converters
Warm Audiences (Retargeting)
Website Retargeting
- All website visitors, last 30 days
- Specific page visitors (pricing, services, case studies), last 60 days
- Add-to-cart / initiate checkout but didn't convert, last 14 days
Engagement Retargeting
- Video viewers (25%, 50%, 75%), last 30 days
- Post/ad engagers, last 30 days
- Instagram profile visitors, last 30 days
- Lead form openers who didn't submit, last 30 days
Audience Sizing
For each audience:
- Estimate audience size based on targeting parameters
- Flag audiences that are too small (< 10K) or too broad (> 10M)
- Recommend adjustments to hit the sweet spot (100K - 2M for prospecting)
Phase 3: Ad Copy Generation
Primary Text (up to 600 characters recommended, 125 for preview)
Write 3-5 variants for each ad set:
Variant Types
- Pain point lead: Open with the audience's biggest frustration
- Result lead: Open with a specific outcome or case study number
- Question lead: Open with a question that triggers self-identification
- Story lead: Open with a mini narrative or client transformation
- Authority lead: Open with a credibility signal or industry insight
Headline (max 40 characters)
Write 3-5 headline options:
- Direct benefit statement
- Offer/lead magnet focused
- Social proof focused
- Curiosity/question focused
- Urgency focused
Description (max 30 characters)
Write 2-3 description options:
- Reinforce the headline
- Add a supporting benefit
- Include a CTA variation
CTA Button
Recommend from Meta's options:
- Learn More (awareness/consideration)
- Sign Up (lead gen)
- Get Quote (services)
- Shop Now (ecommerce)
- Book Now (appointments)
- Download (lead magnet)
- Contact Us (direct response)
Copy Rules
- First line must hook attention (appears in preview)
- Use line breaks for readability
- Include social proof numbers where possible
- Match the ad copy tone to the audience temperature (cold = educational, warm = direct)
- No clickbait, no misleading claims
- Emojis: use sparingly and only if on-brand (max 2-3 per ad)
Phase 4: Creative Format Recommendations
Format Selection by Objective
| Objective | Best Formats | Why |
|---|---|---|
| Awareness | Video (15-30s), Carousel | Video drives recall, carousel tells a story |
| Traffic | Single image, Link ad | Clear CTA, fast load |
| Lead gen | Single image, Carousel, Video | Image for simplicity, carousel for education |
| Conversion | Single image, Video, Collection | Direct response, product showcase |
Creative Specs per Format
Single Image
- Recommended: 1080 x 1080 (1:1) for feed, 1080 x 1920 (9:16) for Stories/Reels
- Max file size: 30MB
- Text on image: < 20% of image area
- Clear visual hierarchy: image > headline > CTA
Carousel (2-10 cards)
- Each card: 1080 x 1080
- Tell a sequential story or showcase multiple benefits/services
- First card must hook, last card must CTA
- Consistent visual style across all cards
Video
- Feed: 1:1 or 4:5, 15-60 seconds
- Stories/Reels: 9:16, 15-30 seconds
- Hook in first 3 seconds
- Captions required (85% of video is watched without sound)
- End with clear CTA frame
UGC Style
- Raw, authentic look outperforms polished in many B2B contexts
- Talking head + screen recording
- Client testimonial clips
- Behind-the-scenes process footage
Creative Brief per Ad Set
For each ad set, specify:
- Visual concept (what the image/video shows)
- Text overlay (if any)
- Mood and style (professional, casual, bold, minimal)
- Reference to
ad-creative-intelfor winning angle inspiration
Phase 5: Budget and Bidding
Budget Allocation Strategy
Distribute monthly budget across ad sets:
| Ad Set | Budget % | Rationale |
|---|---|---|
| Cold - Interest | 35% | Largest audience, prospecting |
| Cold - Lookalike | 25% | High-quality prospecting |
| Warm - Website Retarget | 25% | Highest conversion rate |
| Hot - Engaged Retarget | 15% | Smallest audience, most efficient |
Bidding Strategy
- Lead gen: Lowest cost per result (start), then move to cost cap once CPA is established
- Conversions: Lowest cost, then Target ROAS
- Awareness: Lowest cost per 1000 impressions
- Traffic: Lowest cost per link click
Testing Plan
- Start with 2-3 ad variants per ad set
- Run for 3-5 days before judging performance
- Kill underperformers (CTR < 1% on feed, CPA > 2x target)
- Scale winners by duplicating ad set with 20% budget increase
Phase 6: Output
Return structured JSON:
{
"business_name": "Example Agency",
"built_at": "2024-01-15T14:30:00Z",
"objective": "lead_gen",
"target_market": "India",
"monthly_budget": "30000 INR",
"campaigns": [
{
"campaign_name": "Lead Gen - Shopify Services",
"objective": "Leads",
"optimization_event": "Lead form submission",
"ad_sets": [
{
"ad_set_name": "Cold - Interest Targeting",
"audience": {
"interests": ["Shopify", "Ecommerce", "D2C brands", "Online retail"],
"behaviors": ["Small business owners", "Business page admins"],
"demographics": { "age": "25-50", "location": "India" },
"exclusions": ["Existing customers", "Recent converters"],
"estimated_size": "500K - 1.2M"
},
"budget_allocation": "35%",
"placements": ["Facebook Feed", "Instagram Feed", "Instagram Stories"],
"ads": [
{
"ad_name": "Pain Point - Store Not Converting",
"format": "Single Image",
"primary_text": "Your Shopify store is getting traffic but not sales?\n\nMost D2C brands lose 60-70% of potential revenue to poor store UX.\n\nWe helped Kibi Sports increase conversions by 40% in 8 weeks with our CRO audit process.\n\nBook a free 15-min store audit and see exactly what's costing you sales.",
"headline": "Free Shopify Store Audit",
"description": "Book Your Audit Today",
"cta": "Sign Up",
"landing_page": "https://example.com/free-audit",
"creative_brief": "Image of a Shopify dashboard with conversion metrics highlighted. Clean, professional, data-focused."
}
]
}
]
}
],
"testing_plan": {
"initial_variants": 3,
"evaluation_period": "5 days",
"kill_criteria": "CTR < 0.8% or CPA > 2x target",
"scale_criteria": "CPA < target for 3 consecutive days"
},
"projected_metrics": {
"estimated_reach": "50K-100K per month",
"estimated_cpm": "80-150 INR",
"estimated_ctr": "1-2.5%",
"estimated_cpl": "200-500 INR",
"estimated_monthly_leads": "60-150"
}
}
Example Usage
Trigger phrases:
- "Build Meta ads campaigns for [business]"
- "Create Facebook ad campaigns for [service]"
- "Instagram ads for [product/service]"
- "Meta ads audience targeting for [ICP]"
- "Write Facebook ad copy for [offer]"
User: Build Meta ads for plasho.com -- target D2C brand founders in India for lead gen
Assistant: [reads config, builds interest + lookalike audiences from ICP, generates ad copy variants, recommends creative formats, allocates budget, returns complete campaign blueprint]
User: Write Facebook ad copy for a free Shopify audit offer
Assistant: [generates 5 primary text variants, 5 headlines, 3 descriptions, recommends CTA and creative format, returns formatted ad copy set]
User: Meta ads retargeting campaign for website visitors
Assistant: [builds retargeting audiences by engagement level, writes warm-audience copy, recommends budget allocation, returns campaign structure]
Files bundled with it
These load only when the skill asks for them, so they cost nothing until it runs.
Other skills for the same job
Different authors, same problem. Matched on the words in the skill name, across every library in the catalogue except this one.
- meta-ads-expert by Varnan-Tech · 626
- meta-ads-campaign-builder by edupegoretti · 0
- ads by coreyhaines31 · 46,928
- facebook-ads by OpenClaudia · 677
- google-ads by OpenClaudia · 677
- google-ads-report by OpenClaudia · 677
- linkedin-ads by OpenClaudia · 677
- meta-tribe-skill by Varnan-Tech · 626
Need help setting it up?
This page tells you what meta-ads-builder does and what it needs. Cheetah builds the agent setup it runs inside: data, CRM, sequencing and the guardrails.
Book a call →The directory stays free. There is nothing gated behind this.