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 kokesaurio/mercadolibre-algoritmodigital --skill responder-preguntas-mlgit clone --depth 1 https://github.com/kokesaurio/mercadolibre-algoritmodigitalWrote 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/kokesaurio/mercadolibre-algoritmodigital/responder-preguntas-ml)<a href="https://agentmods.dev/skills/kokesaurio/mercadolibre-algoritmodigital/responder-preguntas-ml"><img src="https://agentmods.dev/badge/skills/kokesaurio/mercadolibre-algoritmodigital/responder-preguntas-ml/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/kokesaurio/mercadolibre-algoritmodigital/responder-preguntas-ml"><img src="https://agentmods.dev/badge/skills/kokesaurio/mercadolibre-algoritmodigital/responder-preguntas-ml.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 | $0.00088 | $0.00618 |
| Opus 5 | $0.00044 | $0.00309 |
| Sonnet 5 | $0.00018 | $0.00124 |
| Haiku 4.5 | $0.00009 | $0.00062 |
Grade A, and why
responder-preguntas-ml 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 3d 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.
What it actually says
Responder preguntas de compradores en MercadoLibre
Flujo para despachar preguntas pendientes rápido y bien: cada hora sin responder baja la chance de venta.
Requisito
Herramientas ml_* del conector de Algoritmo Digital. Si no están, indicá
agregar el conector desde https://mcp.algoritmodigital.com.ar y frená.
Con varias tiendas, resolvé la cuenta con ml_cuentas.
Flujo
- Traer pendientes con
ml_preguntas(por defecto ya filtrapend). Si no hay, avisá y ofrecé revisar las respondidas oml_mensajes. - Contexto antes de redactar: mirá el título y precio de la publicación
(vienen con la pregunta o via
ml_publicacionescon el item_id). Si la pregunta es de stock, verificá conml_stockantes de prometer nada. - Redactar TODAS las respuestas propuestas juntas, numeradas, y
presentarlas al usuario para que apruebe, edite o descarte cada una.4. Publicar solo lo aprobado con
ml_responder_pregunta, de a una. Es una escritura real, pública e irreversible: jamás la ejecutes sin confirmación explícita de ese texto exacto en esta conversación. - Cerrar con el resumen: respondidas, pendientes que quedaron y por qué.
Reglas de redacción (políticas de MercadoLibre)
- Nunca incluir teléfonos, mails, redes, links externos ni invitar a contactar por fuera: MercadoLibre suspende publicaciones por eso.
- Responder solo lo que los datos confirman. Si no hay dato (talle, stock, compatibilidad), la respuesta lo dice honestamente y ofrece alternativa.
- Tono cordial y vendedor: saludo breve, respuesta directa, cierre que invite a comprar ("¡Comprá con confianza!", "Estamos para ayudarte").
- Corta: 1 a 3 oraciones. El comprador lee en el celular.
- Si la pregunta es un reclamo encubierto, no improvisar: sugerir revisar
ml_reclamosy tratarlo por ahí.
Qué no hacer
- No responder en lote sin aprobación individual o de la lista completa.
- No prometer envíos, descuentos ni stock que las herramientas no confirmen.
- No discutir con compradores agresivos: respuesta neutra y profesional.
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.
- 3d ago First seen · 46 lines · 88 tokens per session scan A 5be2c858c735
responder-preguntas-ml is a skill published in the GitHub repository kokesaurio/mercadolibre-algoritmodigital (0 stars, last pushed 4d ago), licensed MIT. It adds 88 tokens to every session and 618 once invoked, about $0.0004 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 2026-09-09.
Other skills, from other repositories
category-opportunity-index
Reports the JoomPulse opportunity level (low, medium, high) for one category on Mercado Livre (Brasil) or Shopee Brasil, with that month's snapshot — estimated revenue and sales, sellers, listings, average ticket — plus concentration and month-over-month growth, and a plain-language verdict on entering it. Use for a…
ml-product-analysis
Analyzes one product and its competitors on JoomPulse — Mercado Livre (Brasil) or Shopee Brasil. From a marketplace link, a JoomPulse link, a photo, or a row of data, it returns a card for the subject product plus a ranked table of comparable items, each with price, estimated monthly sales and revenue, rating and…
seller-copilot
Deeper Mercado Livre (Brasil) and Shopee Brasil seller analyses on JoomPulse data, beyond what a single skill in this repo answers: real margins and fees ("minha margem real", "taxas do ML"); pricing and price wars ("qual preço cobrar", "estou caro?"); is a product worth selling ("vale a pena vender isso"); finding…
seller-overview-tracker
Snapshot tracker for one seller on JoomPulse — Mercado Livre (Brasil) or Shopee Brasil. Pulls store name, estimated monthly revenue and sales, listing count, sales trend, categories, location, and the seller's reputation or buyer rating; medal, cancellation rate and 60/365-day sales are Mercado Livre only. Each run…
uncontested-niche-finder
Finds low-competition niche products in the deep sub-categories of one category on JoomPulse — Mercado Livre (Brasil) or Shopee Brasil. Keeps only the deep niches with no dominant incumbent: no platinum seller on Mercado Livre, no Official store (Shopee Mall) seller on Shopee. Returns one table of niche products, each…
category-monitor
Monitors one category's aggregate health on JoomPulse — Mercado Livre (Brasil) or Shopee Brasil: estimated sales, product count, seller count, the seller-tier mix and market concentration. Each run builds today's snapshot table and offers it for download; send a table from a previous period and the skill shows the…