Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/tray-tecnologia/tray-api-ai-pluginnpx agentmods add skills/tray-tecnologia/tray-api-ai-plugin/cuponsWrote 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/tray-tecnologia/tray-api-ai-plugin/cupons)<a href="https://agentmods.dev/skills/tray-tecnologia/tray-api-ai-plugin/cupons"><img src="https://agentmods.dev/badge/skills/tray-tecnologia/tray-api-ai-plugin/cupons/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/tray-tecnologia/tray-api-ai-plugin/cupons"><img src="https://agentmods.dev/badge/skills/tray-tecnologia/tray-api-ai-plugin/cupons.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 18 findings, 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 Tool Misuse · line 15 Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).Fix: Validate all tool parameters against an allowlist. Reject dangerous parameter values (shell=True, --force, -rf /) and use safe defaults.
- high Tool Misuse · line 93 Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).Fix: Validate all tool parameters against an allowlist. Reject dangerous parameter values (shell=True, --force, -rf /) and use safe defaults.
- high Tool Misuse · line 434 Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).Fix: Validate all tool parameters against an allowlist. Reject dangerous parameter values (shell=True, --force, -rf /) and use safe defaults.
- medium Agent Snooping · line 45 Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.Fix: Remove all code or instructions that list or read other skills' files or directories. Skills should operate independently; cross-skill access is a privilege escalation.
- medium Server-Side Request Forgery · line 188 Request target host is built from a dynamic or untrusted value. If the host is attacker-influenced, this enables SSRF to arbitrary internal or metadata endpoints.Fix: Do not build request URLs from untrusted input. Validate the host against an allowlist and reject internal/metadata addresses before issuing the request.
- medium Server-Side Request Forgery · line 240 Request target host is built from a dynamic or untrusted value. If the host is attacker-influenced, this enables SSRF to arbitrary internal or metadata endpoints.Fix: Do not build request URLs from untrusted input. Validate the host against an allowlist and reject internal/metadata addresses before issuing the request.
- medium Server-Side Request Forgery · line 335 Request target host is built from a dynamic or untrusted value. If the host is attacker-influenced, this enables SSRF to arbitrary internal or metadata endpoints.Fix: Do not build request URLs from untrusted input. Validate the host against an allowlist and reject internal/metadata addresses before issuing the request.
- medium Server-Side Request Forgery · line 405 Request target host is built from a dynamic or untrusted value. If the host is attacker-influenced, this enables SSRF to arbitrary internal or metadata endpoints.Fix: Do not build request URLs from untrusted input. Validate the host against an allowlist and reject internal/metadata addresses before issuing the request.
- medium Server-Side Request Forgery · line 461 Request target host is built from a dynamic or untrusted value. If the host is attacker-influenced, this enables SSRF to arbitrary internal or metadata endpoints.Fix: Do not build request URLs from untrusted input. Validate the host against an allowlist and reject internal/metadata addresses before issuing the request.
- medium Server-Side Request Forgery · line 545 Request target host is built from a dynamic or untrusted value. If the host is attacker-influenced, this enables SSRF to arbitrary internal or metadata endpoints.Fix: Do not build request URLs from untrusted input. Validate the host against an allowlist and reject internal/metadata addresses before issuing the request.
- medium Server-Side Request Forgery · line 577 Request target host is built from a dynamic or untrusted value. If the host is attacker-influenced, this enables SSRF to arbitrary internal or metadata endpoints.Fix: Do not build request URLs from untrusted input. Validate the host against an allowlist and reject internal/metadata addresses before issuing the request.
- medium Server-Side Request Forgery · line 607 Request target host is built from a dynamic or untrusted value. If the host is attacker-influenced, this enables SSRF to arbitrary internal or metadata endpoints.Fix: Do not build request URLs from untrusted input. Validate the host against an allowlist and reject internal/metadata addresses before issuing the request.
- medium Server-Side Request Forgery · line 637 Request target host is built from a dynamic or untrusted value. If the host is attacker-influenced, this enables SSRF to arbitrary internal or metadata endpoints.Fix: Do not build request URLs from untrusted input. Validate the host against an allowlist and reject internal/metadata addresses before issuing the request.
- medium Server-Side Request Forgery · line 667 Request target host is built from a dynamic or untrusted value. If the host is attacker-influenced, this enables SSRF to arbitrary internal or metadata endpoints.Fix: Do not build request URLs from untrusted input. Validate the host against an allowlist and reject internal/metadata addresses before issuing the request.
- medium Server-Side Request Forgery · line 700 Request target host is built from a dynamic or untrusted value. If the host is attacker-influenced, this enables SSRF to arbitrary internal or metadata endpoints.Fix: Do not build request URLs from untrusted input. Validate the host against an allowlist and reject internal/metadata addresses before issuing the request.
- medium Server-Side Request Forgery · line 754 Request target host is built from a dynamic or untrusted value. If the host is attacker-influenced, this enables SSRF to arbitrary internal or metadata endpoints.Fix: Do not build request URLs from untrusted input. Validate the host against an allowlist and reject internal/metadata addresses before issuing the request.
- medium Data Exfiltration · line 732 Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
- medium Agent Snooping · line 965 Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.Fix: Remove all code or instructions that list or read other skills' files or directories. Skills should operate independently; cross-skill access is a privilege escalation.
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 | $0.00156 | $0.16551 |
| Opus 5 | $0.00078 | $0.08275 |
| Sonnet 5 | $0.00031 | $0.03310 |
| Haiku 4.5 | $0.00016 | $0.01655 |
Grade A, and why
tray-cupons scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
- **Exemplo (curl):** How it starts
The opening of the file, as written. The whole thing — 1,010 lines — stays where its author put it; the contents beside it link to each section on GitHub.
MANDATORY: Tool Calls Required Before Answering
Estas chamadas são OBRIGATÓRIAS, não opcionais. Execute-as antes de gerar qualquer código ou payload. Se você está respondendo sem ter chamado a ferramenta abaixo, pare e chame agora.
1. Buscar documentação atualizada (sempre)
node skills/tray-dev/scripts/search_docs.mjs --topic=cupons "<termo da pergunta>"
<TOPIC_SLUG>: ver tabela emskills/tray-dev/SKILL.md.- Use os trechos retornados como fonte primária; este SKILL.md é resumo denso.
2. Revisar campos (este recurso ainda NÃO tem validate.mjs)
Nota: o recurso
cuponsainda não possuiscripts/validate.mjslocal. A chamada OBRIGATÓRIA asearch_docs.mjsacima continua valendo. Como não há validador automático, você é responsável por revisar manualmente cada campo obrigatório (code,value,type) contra a doc retornada porsearch_docs.mjse contra os schemas de referência emskills/cupons/schemas/antes de retornar qualquer código. Confira em especial: chave de recursoDiscountCouponpresente no body,codesem espaços nem acentos,typeem$/%, e coerência entreusage_counter_limiteusage_counter_limit_customer.
Antes de responder
Execute estas verificações antes de gerar qualquer payload ou código:
- Confirme o método HTTP e o endpoint correto para a operação solicitada
(CRUD em
/discount_coupons, consulta de relacionamento em/discount_coupons/<tipo>_relationship/:id, ou criação de vínculo em/discount_coupons/create_relationship/:id). - Identifique os campos obrigatórios listados neste documento —
code,valueetypenunca podem faltar na criação; não omita nenhum. - Verifique que
access_tokennão aparece como literal string no código gerado — use sempreTRAY_ACCESS_TOKENeTRAY_API_ADDRESSpor variável de ambiente. - Confirme que esta é a skill correta para o recurso: cupom é desconto por
código no checkout; se for preço permanente por grupo, leia
when_not_to_usee redirecione paratray-listas-preco-b2b.
What ships with it
29 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- examples/cupom-atualizar.curl.sh 1.1 KB runs code
- examples/cupom-atualizar.fixture.json 144 B
- examples/cupom-atualizar.node.mjs 1.5 KB runs code
- examples/cupom-consultar.curl.sh 625 B runs code
- examples/cupom-consultar.node.mjs 889 B runs code
- examples/cupom-criar-relacionamento.curl.sh 967 B runs code
- examples/cupom-criar-relacionamento.fixture.json 94 B
- examples/cupom-criar-relacionamento.node.mjs 1.4 KB runs code
- examples/cupom-criar.curl.sh 1.1 KB runs code
- examples/cupom-criar.fixture.json 289 B
- examples/cupom-criar.node.mjs 1.4 KB runs code
- examples/cupom-excluir.curl.sh 856 B runs code
- examples/cupom-excluir.node.mjs 1.1 KB runs code
- examples/cupom-listar.curl.sh 568 B runs code
- examples/cupom-listar.node.mjs 854 B runs code
- examples/cupom-relacionamento-categorias.curl.sh 655 B runs code
- examples/cupom-relacionamento-categorias.node.mjs 935 B runs code
- examples/cupom-relacionamento-clientes.curl.sh 667 B runs code
- examples/cupom-relacionamento-clientes.node.mjs 945 B runs code
- examples/cupom-relacionamento-fretes.curl.sh 667 B runs code
- examples/cupom-relacionamento-fretes.node.mjs 943 B runs code
- examples/cupom-relacionamento-marcas.curl.sh 644 B runs code
- examples/cupom-relacionamento-marcas.node.mjs 920 B runs code
- examples/cupom-relacionamento-presente.curl.sh 667 B runs code
- examples/cupom-relacionamento-presente.node.mjs 945 B runs code
- examples/cupom-relacionamento-produtos.curl.sh 679 B runs code
- examples/cupom-relacionamento-produtos.node.mjs 957 B runs code
- schemas/discount_coupons.create.json 3.5 KB
- schemas/discount_coupons.update.json 3.0 KB
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.
- 9d ago First seen · 1,010 lines · 156 tokens per session scan A 07eb3444360f
tray-cupons is a skill published in the GitHub repository tray-tecnologia/tray-api-ai-plugin (15 stars, last pushed 27d ago), licensed MIT. It adds 156 tokens to every session and 16,551 once invoked, about $0.0008 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
ontopo
Find restaurants and check table availability in Israel. Search across multiple dates and venues at once, view menus, and get a booking link to finish on Ontopo. Covers dining and restaurant discovery in Tel Aviv, Jerusalem, Haifa, Eilat and 23 more cities. Use for "restaurant reservation", "table booking", "book a…
buywhere-product-catalog
Use BuyWhere's MCP and API surfaces to add product search, price comparison, and deal discovery to AI shopping agents.
buywhere-product-catalog
Use BuyWhere's MCP and API surfaces to add product search, price comparison, and deal discovery to AI shopping agents.
etsy-seo
Etsy shop SEO optimization toolkit. This skill should be used when the user asks to "analyze my Etsy shop", "optimize my Etsy listings", "check my Etsy SEO", "audit Etsy tags", "improve my Etsy sales", "compare Etsy competitors", "update Etsy listing", or mentions anything related to Etsy shop optimization, listing…
etsy-audit
Full SEO audit of an Etsy shop. Scores every active listing on title, tags, description, and images (0-100). Identifies worst performers and provides prioritized recommendations. Use when the user says "audit my Etsy shop", "check my Etsy SEO scores", "how are my listings performing", or "which listings need work".
etsy-competitors
Search the Etsy marketplace for competing listings and analyze their SEO strategies. Identifies keyword gaps, pricing patterns, and title strategies. Use when the user says "find competitors", "competitor analysis", "what are others doing", "keyword gaps", or "how do I compare to competitors".