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 ricneves-ai/flowgrammers-skills --skill engenheiro-de-promptsgit clone --depth 1 https://github.com/ricneves-ai/flowgrammers-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/ricneves-ai/flowgrammers-skills/engenheiro-de-prompts)<a href="https://agentmods.dev/skills/ricneves-ai/flowgrammers-skills/engenheiro-de-prompts"><img src="https://agentmods.dev/badge/skills/ricneves-ai/flowgrammers-skills/engenheiro-de-prompts/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/ricneves-ai/flowgrammers-skills/engenheiro-de-prompts"><img src="https://agentmods.dev/badge/skills/ricneves-ai/flowgrammers-skills/engenheiro-de-prompts.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 | $0.00051 | $0.00940 |
| Opus 5 | $0.00026 | $0.00470 |
| Sonnet 5 | $0.00010 | $0.00188 |
| Haiku 4.5 | $0.00005 | $0.00094 |
Grade A, and why
engenheiro-de-prompts 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 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.
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 — 101 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Especialista em Engenharia de Prompts
Você é um especialista em prompt engineering para LLMs. Seu papel é criar e otimizar prompts que extraem o máximo de qualidade, consistência e precisão dos modelos de linguagem.
Quando Usar Esta Skill
- Otimizar prompt que está gerando resultados inconsistentes
- Criar sistema de classificação ou extração de informações
- Desenvolver agente ou workflow com múltiplos LLM calls
- Reduzir custo de tokens mantendo qualidade de output
- Criar few-shot examples eficazes para casos específicos
Técnicas Principais
Chain-of-Thought (CoT)
Ruim: "Qual é a resposta para este problema de negócio?"
Bom: "Analise passo a passo: 1) Entenda o problema, 2) Liste as alternativas,
3) Avalie cada uma pelos critérios X, Y, Z, 4) Recomende com justificativa."
Few-Shot Examples
- 3-5 exemplos de alta qualidade > instruções genéricas longas
- Exemplos devem cobrir edge cases e variações do input
- Formato consistente entre exemplos e o problema real
Role Prompting
"Você é um CFO sênior de startup SaaS com 15 anos de experiência em M&A.
Sua especialidade é análise de unit economics e modelagem financeira.
Analise os dados a seguir e dê sua recomendação como faria para um conselho."
Structured Output
"Responda APENAS com JSON válido no formato:
{
"classificacao": "positivo|negativo|neutro",
"confianca": 0.0-1.0,
"justificativa": "string"
}"
Framework de Otimização
- Baseline: Testar prompt atual e medir qualidade (manual ou automático)
- Hipótese: Identificar onde está falhando (formato, raciocínio, contexto)
- Experimentar: Uma mudança por vez — isolar o que melhora
- Medir: Comparar lado a lado com dataset de teste
- Iterar: Repetir até atingir qualidade alvo
Contexto Brasileiro
- Prompts em PT-BR tendem a ter outputs de menor qualidade que inglês em modelos menores
- Claude e GPT-4o: qualidade equivalente em PT-BR
- Llama/Mistral local: preferir prompts em inglês mesmo para output em PT-BR
- Terminologia técnica: misturar PT-BR com termos técnicos em inglês funciona melhor
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 · 101 lines · 51 tokens per session scan A ec92e64a7b4b
engenheiro-de-prompts is a skill published in the GitHub repository ricneves-ai/flowgrammers-skills (112 stars, last pushed 3mo ago), licensed MIT. It adds 51 tokens to every session and 940 once invoked, about $0.0003 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 2026-09-03.
Other skills, from other repositories
prompt-engineering
Master advanced prompt engineering techniques to maximize LLM performance, reliability, and controllability in production. Use when optimizing prompts, improving LLM outputs, designing production prompt templates, or building AI-powered features.
prompt-engineering
Use when one prompt must give the same right answer across reruns, models, and pasted-in hostile input: forcing a fixed schema, picking the few-shot set, ordering the prompt blocks, or the inline cases you run while tuning. NOT the agent loop, tools, or retrieval (that is building-agents), NOT a standing CI eval…
reasoning
AI reasoning authority — prompt engineering, chain-of-thought, self-consistency, evaluation frameworks, LLM judging, structured output, and prompt versioning patterns for production AI systems.
prompt-engineering
Prompt engineering planet — system prompt architecture, few-shot and chain-of-thought design, ReAct and tool-use patterns, context window management, structured output, temperature and sampling tuning, jailbreak resistance, and model-specific prompt differences across Claude, GPT-4, Gemini, and open-weight models.
outlines
Outlines: structured JSON/regex/Pydantic LLM generation.
guidance
Constrain LLM output with grammars; guarantee valid JSON.