Ai 마법서

Template-05-Tag

3 min read

<%* // Gemini API 키 설정 const GEMINI_API_KEY=“AIzaSyAUyWPGBkl9fxpAh1O4uIGU87I2dpgSYOg”%>

<%_* // 태그 생성을 위한 프롬프트 정의 const tag_prompt = `You are an expert at generating precise, SEO-optimized tags for technical and business documentation. Your task is to analyze the given content and create highly searchable tags that capture core concepts and technical elements.

Analysis Framework:

  1. Technical Classification:
  • Primary Technologies (프레임워크/라이브러리/도구)
  • Technical Concepts (알고리즘/아키텍처/패턴)
  • Implementation Details (구현방식/메서드/기술스택)
  • Domain Specifics (분야/산업/적용영역)
  1. Contextual Analysis:
  • Business Value (ROI/성과/효율)
  • Use Cases (활용사례/적용방안)
  • Problem Domains (해결과제/니즈)
  • Target Users (사용자/클라이언트)
  1. Tag Selection Criteria: Must Meet ALL Conditions:
  • Explicitly appears in content
  • Highly specific (no general terms)
  • Technical precision
  • Search optimization value
  • Clear relation to core topic

Tag Generation Rules:

Technical Requirements:

  • Keep technical terms in English (#RAG, LLM etc)
  • Include version numbers when specific
  • Use established technical terminology
  • Maintain technical accuracy

Tag Structure:

  • 1-10 tags only
  • NO spaces within tags
  • Korean preferred except for technical terms
  • Compound words allowed for precision
  • Follow established tech community conventions

Exclusion Rules:

  • NO general/vague terms
  • NO marketing buzzwords
  • NO compound tags mixing Korean/English
  • NO tags not present in content
  • NO subjective descriptors

Quality Checks:

  • Verify each tag appears in content
  • Ensure searchability
  • Check technical accuracy
  • Validate specificity
  • Confirm relevance

[Output Requirements]

  • ONLY comma-separated tags
  • Each tag starts with #
  • NO explanations or alternatives
  • Single line output
  • NO markdown formatting

Example format: #RAG, 문서처리, 파이프라인, LangChain, 성능최적화

Remember:

  • Tags must be precise, present in content, and highly searchable while maintaining technical accuracy.
  • Generate maximum 10 tags. ` _%>

<%_* // frontmatter의 tags 속성을 업데이트하는 함수 const processTags = async (file, newTags) ⇒ { await tp.app.fileManager.processFrontMatter(file, (frontmatter) ⇒ { const tags = newTags.split(’,‘).map(tag ⇒ tag.trim()); frontmatter.tags = tags.map(tag ⇒ ${tag}); }); }; _%>

<%_* // Gemini API를 호출하여 태그를 생성하는 함수 async function generateTags(content) { const model = “gemini-1.5-flash”; const url = https://generativelanguage.googleapis.com/v1beta/models/${model}:generateContent?key=${GEMINI_API_KEY}; const full_prompt = ${tag_prompt}\n\nHere is the content of the note:\n${content};

try {
    const response = await tp.obsidian.requestUrl({
        method: "POST",
        url: url,
        contentType: "application/json",
        body: JSON.stringify({
            contents: [{ parts: [{ text: full_prompt }] }]
        })
    });
    
    const resultText = response.json.candidates[0].content.parts[0].text;
    return resultText.trim();
} catch (error)
{
    console.error('Gemini API 호출 중 오류 발생:', error);
    if (error.response) {
        console.error('Gemini API Error Response:', await error.response.text());
    }
    return '태그 생성 실패';
}

}

// 현재 노트의 내용을 가져옵니다 const fileContent = tp.file.content;

// 태그를 생성하고 파일에 적용 const tags = await generateTags(fileContent); const file = tp.config.target_file; await processTags(file, tags); _%>

Graph View