Skip to main content

prompting

Guide for writing effective system prompts for LLM agents. Use when creating or editing system prompts for applications, agent configurations, or development tools.

Ir para a instalação

Informações da origem

Repositório
cashew-labs/libretto
Última atividade na origem
4 de março de 2026 às 23:18
Idioma detectado do SKILL.md
inglês
Estrelas
899
Forks
69

Opções de instalação

Por padrão, está selecionado o prompt que primeiro revisa a origem. Você pode mudar para um comando direto ou baixar uma cópia local.

Revise os arquivos de origem

Leia o SKILL.md e os arquivos complementares exibidos pelo SkillsMP antes de decidir se vai instalar.

Explorador de arquivos
5 arquivos

Exibindo SKILL.md

SKILL.md
Instruções da origem · Visualização somente leitura
name
prompting
description
Guide for writing effective system prompts for LLM agents. Use when creating or editing system prompts for applications, agent configurations, or development tools.
# Prompting ## Philosophy LLMs are intelligent by default. System prompts set direction and impose constraints, not explain reasoning. Start minimal. Observe failures. Add targeted fixes. Every instruction must justify its token cost by solving a real problem. Do not explain existing capabilities, list obvious practices, add preventive instructions, or repeat information. ## Structure Use markdown sections and paragraphs. Each section describes one behavior or constraint. State what to do or avoid. Explain why if non-obvious. Show correct behavior with examples. ### Formatting Headings up to level 3. Plain paragraphs. No bold, italics, or emojis. Code blocks for commands. Lists only for distinct enumerable items. ## Examples Wrap examples in `<example>` tags with user/assistant prefixes. One pair per tag. ``` <example> user: What's the capital of France? assistant: Paris </example> ``` Use brackets for tool actions instead of showing invocations: ``` <example> user: Find all TODO comments assistant: [searches codebase] Found 3 TODOs: ... </example> ``` ## Include Behaviors the model gets wrong by default. Domain constraints. Output format requirements. Safety boundaries. Tool integrations. ## Omit Reasoning instructions. Problem-solving approaches. Common sense behaviors. Ethical guidelines. Capability descriptions. ## Iteration Start minimal. Test with real inputs. Identify failures. Add targeted fixes. Remove unnecessary instructions. Track which instructions prevent which failures. If you cannot identify the specific problem an instruction solves, remove it. ## Model-Specific Guidance Consult references/ for model-specific patterns: - references/claude.md - XML structure, countering sycophancy, trigger words, parallel execution - references/gpt.md - Contradiction sensitivity, role hierarchy, verbosity control, metaprompting - references/gemini.md - Conciseness, tool explanations, library checks, context placement - references/codex.md - OpenAI Codex models, tool implementations, autonomy patterns, compaction
Ver no GitHub