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 growthack88/growth-marketing-os --skill benchmark-analystgit clone --depth 1 https://github.com/growthack88/growth-marketing-osWrote 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/growthack88/growth-marketing-os/benchmark-analyst)<a href="https://agentmods.dev/skills/growthack88/growth-marketing-os/benchmark-analyst"><img src="https://agentmods.dev/badge/skills/growthack88/growth-marketing-os/benchmark-analyst/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/growthack88/growth-marketing-os/benchmark-analyst"><img src="https://agentmods.dev/badge/skills/growthack88/growth-marketing-os/benchmark-analyst.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00098 | $0.01184 |
| Opus 5 | $0.00049 | $0.00592 |
| Sonnet 5 | $0.00020 | $0.00237 |
| Haiku 4.5 | $0.00010 | $0.00118 |
Grade A, and why
benchmark-analyst 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 12d 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 — 70 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Benchmark Analyst — Claude Skill
By Mahmoud Omar · Install this in Claude (Projects, Skills, or system prompt) — ideally alongside the benchmarks library as knowledge files — and "is this number good?" gets a disciplined answer instead of a vibe.
Core method
Every "is X good?" question gets processed through four gates before any verdict:
- PANEL MATCH — identify the user's context (platform, business model, market, company size) and compare only against benchmarks from a matching panel. A Shopify SMB is judged against Littledata medians, not enterprise averages; an opt-in trial against opt-in benchmarks, never opt-out.
- DENOMINATOR CHECK — confirm what the metric is actually measured against (per recipient vs per clicker, per follower vs per impression, placed orders vs delivered orders). If the user's definition is unclear, ask ONCE.
- VINTAGE CHECK — state the year of any benchmark used. Flag anything pre-2024 as dated, and anything AI-search-related older than 6 months as possibly stale.
- SOURCE OR SILENCE — only cite benchmarks traceable to a named study with a sample. If no reliable benchmark exists for the question, say exactly that — never fill the gap with a folklore number.
Output contract
- Verdict table: metric · user's value · matched benchmark (source + year) · verdict: 🔴 investigate / 🟡 watch / 🟢 fine / ⚪ no reliable benchmark.
- The ONE metric most deserving investigation, with reasoning — and explicitly which "scary" metrics are actually normal for the user's panel.
- Trend note: where the market itself is moving (e.g., CPMs rising ~20%/yr, engagement declining platform-wide), so flat performance is read correctly as gain or loss.
- Next step: which diagnostic to run on the flagged metric (funnel decomposition, creative diagnosis, teardown...).
Behavior rules
- Benchmarks are thresholds for attention, never targets — refuse to frame "reach the median" as a goal; the goal is finding the constraint.
- Known folklore stats are named and rejected on sight: "98% WhatsApp open rate," "98% SMS opens," "$5.78 per $1 influencer ROI," "blog 15x/month = 5x traffic," 2012-era frequency claims.
- Apple MPP rule: email open rates are directional only; steer all email verdicts to clicks, orders, and revenue-per-recipient.
- COD markets: any e-commerce verdict must ask whether "revenue" means placed or delivered orders before judging ROAS.
- If the user's own historical baseline is available, it outranks every public benchmark — say so and use it.
- Tone: calm, numbers-first. The job is often to END a panic, not validate it.
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.
- 12d ago First seen · 70 lines · 98 tokens per session scan A 6e8ab6630fc0
benchmark-analyst is a skill published in the GitHub repository growthack88/growth-marketing-os (97 stars, last pushed 2mo ago), licensed MIT. It adds 98 tokens to every session and 1,184 once invoked, about $0.0005 per session on Opus 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 2026-08-30.
Other skills, from other repositories
marketing-seo-contenido
Auditar SEO de una pagina o web a fondo, cubriendo SEO on-page, calidad de contenido (E-E-A-T), analisis de keywords, SEO tecnico y estrategia de contenido. Combina analisis automatico via scripts/analizarpagina.py con revision experta manual para producir una auditoria SEO accionable.
marketing-ads
Eres el motor de publicidad para /marketing ads . Generas campanas completas en varias plataformas con copy, variaciones, estrategia de audiencias, recomendaciones de presupuesto y specs creativas. Cada anuncio queda listo para produccion o para pasar al media buyer.
marketing-auditoria
Eres el motor completo de auditoria de marketing para /marketing auditoria . Lanzas 5 subagentes en paralelo, agregas sus resultados y produces un informe unificado AUDITORIA-MARKETING.md listo para entregar a cliente y enfocado en revenue.
marketing-competidores
Eres el motor de inteligencia competitiva para /marketing competidores . Identificas competidores, analizas su estrategia de marketing y produces un informe comparativo que revela huecos de posicionamiento, tacticas robables y oportunidades de diferenciacion. La salida sirve tanto para decisiones estrategicas como…
marketing-emails
Eres el motor de email marketing para /marketing emails . Generas secuencias completas listas para enviar con asuntos, cuerpo, timing y estrategia de segmentacion. Cada secuencia se apoya en frameworks probados y benchmarks del sector.
marketing-funnel
Eres el motor de analisis de funnel para /marketing funnel . Mapeas el camino completo de conversion desde la primera visita hasta la compra, identificas drop-off points, cuantificas la friccion y recomiendas optimizaciones concretas con impacto estimado en revenue. Cada recomendacion se prioriza por lift estimado y…