Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add swan-gtm/gtm-skills --skill meta-ads-operating-systemgit clone --depth 1 https://github.com/swan-gtm/gtm-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/swan-gtm/gtm-skills/meta-ads-operating-system)<a href="https://agentmods.dev/skills/swan-gtm/gtm-skills/meta-ads-operating-system"><img src="https://agentmods.dev/badge/skills/swan-gtm/gtm-skills/meta-ads-operating-system/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/swan-gtm/gtm-skills/meta-ads-operating-system"><img src="https://agentmods.dev/badge/skills/swan-gtm/gtm-skills/meta-ads-operating-system.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high Memory Poisoning · line 258 Skill manipulates agent memory, state, or stored context. Memory corruption can alter personality, override safety rules, or cause unpredictable behavior.Fix: Protect agent memory and state from modification by untrusted content. Use read-only memory for critical instructions and validate all state changes.
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00042 | $0.06337 |
| Opus 5.5 | $0.00017 | $0.02535 |
| Sonnet 5.5 | $0.00008 | $0.01267 |
| Haiku 4.5 | $0.00004 | $0.00634 |
Grade A, and why
meta-ads-operating-system scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 469 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Meta Ads Operating System - B2B SaaS
The single decision framework for running Meta ad accounts for B2B SaaS. Every decision - when to swap an ad, when to graduate it, when to scale budget, how many creatives to produce - flows from this system.
This file drives all operational decisions. Other knowledge-base files provide deeper context on specific topics. When in doubt, follow this OS.
Core Principle
Meta's algorithm is excellent at optimizing delivery (getting people to click and submit forms). But it cannot see lead quality. In B2B, a large percentage of form submissions come from people outside the ICP. The algorithm treats all leads as equal. Our job is to add the quality layer Meta cannot see, and make every decision based on qualified leads - not raw form fills.
1. Set the Target
Every formula depends on one number: TCPL (Target Cost Per Lead) = target cost per qualified lead. All thresholds are derived from TCPL. Without TCPL, you cannot run the Decision Tree, classify ads, or make scaling decisions. Establishing TCPL is always Step 1.
Scenario A: You have a target cost per demo (ideal)
TCPL = Target Cost per Demo x QL-to-Demo Rate
Example: Target cost per demo = EUR 2,000. QL-to-demo rate = 28%. TCPL = EUR 2,000 x 0.28 = EUR 560.
This is the strongest TCPL because it connects ad spend directly to a business outcome. Always pursue this number. If you know your target cost per demo but not the QL-to-demo rate, establishing that rate becomes the first measurement priority.
Scenario B: No target provided, but account has historical data
TCPL = 30-day trailing CPL(QL) x 0.80
Timeline: achieve TCPL within 30 days of engagement start
Logic: The 30-day trailing average represents current performance including waste. A 20% reduction is achievable through operational improvements (cutting zero-QL ads, graduating winners, improving creative mix) without structural changes like new audiences or conversion architecture. It is a realistic first target.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 469 lines · 42 tokens per session scan A 513bb88a96f6
meta-ads-operating-system is a skill published in the GitHub repository swan-gtm/gtm-skills (221 stars, last pushed 8d ago), licensed MIT. It adds 42 tokens to every session and 6,337 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other skills, from other repositories
ad-campaign-analyzer
Analyze cross-channel campaign data, quantify uncertainty, and propose evidence-labeled budget tests without overstating causality.
paid-ad-copy-writer
Write paid ad copy for affiliate offers across ad platforms. Triggers on: "write ad copy", "Facebook ad for affiliate", "Google Ads copy", "TikTok ad script", "Pinterest ad", "paid traffic to affiliate", "create ad campaign", "ad headlines", "ad descriptions", "scale with paid ads", "run ads for my affiliate link"…
gooseworks
GooseWorks growth coworker and router for markets, competitors, audiences, creators, campaigns, ads, photos, graphics, video, content, email, SEO, social, leads and contacts. Use for brand facts or copy too, so offers, prices, rules, audience and bestsellers reach the brand's Marketing Brain.
adcp-advertising
Automate advertising campaigns with AI. Create ads, buy media, manage ad budgets, discover ad inventory, run display ads, video ads, CTV campaigns, and optimize ad performance. Perfect for marketing automation, programmatic advertising, media buying, ad management, campaign optimization, creative management, and…
yandex-direct
Skill "yandex-direct" from VKirill/claude-lane-stack, covering version requirements, usage, use this skill when, do not use this skill when and purpose.
yandex-direct-creatives
An advertising-writing guide for Yandex.Direct, a Russian online advertising platform. It turns keyword clusters from Key Collector, a keyword research tool, into campaigns, ad groups, and ads.