文本提及网络访问
6 次命中此源文件行包含 URL 或网络客户端。应检查目标地址、发送的数据和信任边界;静态文本无法证明实际发生了网络活动。 该规则在收集文本中命中 6 处;此处列出 3 个代表位置。命中文本可能只是文档,分析器没有执行它。
技能研究档案
Writes, refactors, and evaluates prompts for LLMs — generating optimized prompt templates, structured output schemas, evaluation rubrics, and test suites. Use when designing prompts for new LLM applications, refactoring existing prompts for better accuracy or token efficiency, implementing chain-of-thought or few-shot learning, creating system prompts with personas and guardrails, building JSON/function-calling schemas, or developing prompt evaluation frameworks to measure and improve model performance.
Writes, refactors, and evaluates prompts for LLMs — generating optimized prompt templates, structured output schemas, evaluation rubrics, and test suites. Use when designing prompts for new LLM applications, refactoring existing prompts for better accuracy or token efficiency, implementing chain-of-thought or few-shot learning, creating system prompts with personas and guardrails, building JSON/function-calling schemas, or developing prompt evaluation frameworks to measure and improve model performance.
Skill Atlas editorial review · editorial-preview-1.0.0
指导 LLM 应用的提示词设计、测试与迭代,涵盖少样本与思维链模式、结构化输出 schema 和评测方案。
对所收集的 SKILL.md 及辅助文本/代码进行确定性静态分析。机器辅助候选和发现等待人工编辑复核;不评估动态行为或效果。
文件清单来自记录提交中的采集源码。路径与大小属于结构观察,不代表质量评价。
skills/prompt-engineer/references/context-management.md引用7.4 KBskills/prompt-engineer/references/evaluation-frameworks.md引用23.2 KBskills/prompt-engineer/references/prompt-optimization.md引用17.2 KBskills/prompt-engineer/references/prompt-patterns.md引用15.8 KBskills/prompt-engineer/references/structured-outputs.md引用18.6 KBskills/prompt-engineer/references/system-prompts.md引用18.5 KBskills/prompt-engineer/SKILL.md指令5.9 KB{
"name": "prompt-engineer",
"description": "Writes, refactors, and evaluates prompts for LLMs — generating optimized prompt templates, structured output schemas, evaluation rubrics, and test suites. Use when designing prompts for new LLM applications, refactoring existing prompts for better accuracy or token efficiency, implementing chain-of-thought or few-shot learning, creating system prompts with personas and guardrails, building JSON/function-calling schemas, or developing prompt evaluation frameworks to measure and improve model performance.",
"license": "MIT",
"metadata": {
"author": "https://github.com/Jeffallan",
"company": "https://synergetic.solutions",
"version": "1.2.0",
"domain": "data-ml",
"triggers": "prompt engineering, prompt optimization, chain-of-thought, few-shot learning, prompt testing, LLM prompts, prompt evaluation, system prompts, structured outputs, prompt design, context management, lost-in-the-middle, context degradation, token optimization, attention budget",
"role": "expert",
"scope": "design",
"output-format": "document",
"related-skills": "test-master, rag-architect, debugging-wizard"
}
}此工作流根据源码材料重建表面顺序。除非明确记录受控评估,否则它属于暂定解读,而非真实运行轨迹。
选择零样本、少样本或思维链并撰写清晰指令。
若测试集准确率低于 80%,先找出失败模式再继续迭代。
提供最终提示词、测试用例与结果、使用设置、指标和已知限制。
静态发现用于提示安装前需要检查的位置,不能证明一项技能安全、恶意或有效。
收集的文本中有 86 项确定性词法发现需要复核。发现可能仅位于文档中,不证明已执行或可被利用;仅含二进制的文件未扫描。
此源文件行包含 URL 或网络客户端。应检查目标地址、发送的数据和信任边界;静态文本无法证明实际发生了网络活动。 该规则在收集文本中命中 6 处;此处列出 3 个代表位置。命中文本可能只是文档,分析器没有执行它。
此行包含可能的文件写入或重定向操作。应检查目标路径和覆盖行为;静态审查没有写入任何文件。 该规则在收集文本中命中 76 处;此处列出 3 个代表位置。命中文本可能只是文档,分析器没有执行它。
所提供的内容包含指令覆盖语言。应将其视为不可信文本,评估它是否试图取代更高优先级指令或泄露隐藏上下文。 该规则在收集文本中命中 4 处;此处列出 3 个代表位置。命中文本可能只是文档,分析器没有执行它。
确认提供给它的提示词和示例不含敏感数据,Skill 本身也明确禁止硬编码敏感信息。
发现、激活与资源层具有不同含义。只有使用指定官方方法并记录模型的运行时,才显示精确数值。
在管线使用明确模型执行官方精确计数之前,包括文件清单在内的所有 token 值均为 null。
未使用要求的官方方法记录 token 测量。本档案不会用字符数、文件大小或估计值替代。
该技能仅具备以上标明的证据。不声称任何任务成功率、token 效率、延迟或质量优势。
git clone https://github.com/Jeffallan/claude-skills.git
cd claude-skills
# 将该 Skill 目录复制到以下任一位置:
# ~/.claude/skills/prompt-engineer
# .claude/skills/prompt-engineer确认目标位置能够发现 prompt-engineer 的 SKILL.md,然后让 Claude Code 在一个小型测试任务中使用它。
Use the prompt-engineer skill to refactor my customer-support ticket classifier prompt. Define success criteria, propose a few-shot version, and write a test set including edge cases such as empty and multilingual tickets.
Read the prompt-engineer skill files at https://github.com/Jeffallan/claude-skills/blob/67c1961b180dc92cf08983344627109e983e8f51/skills/prompt-engineer/SKILL.md. Explain its workflow, required tools, permissions and the main risks before I decide whether to install it.
研究结论一份严谨的提示词工程清单,效果取决于你提供的测试用例。
档案是对公开工件的限时阅读,而非背书。请通过来源链接检查确切的上游材料。
67c1961b180dc92cf08983344627109e983e8f51仅对提供的文件进行确定性词法解析、YAML frontmatter 校验、文件清单、来源描述的工作流程提取和启发式暂定分类。由机器生成;无 LLM、执行、动态测试或人工编辑批准。