jd-analyze

name	jd-analyze
description	Analyze and categorize job description requirements. Parses JD text into structured JobRequirements with categorized requirements (must_have, nice_to_have, culture, logistics), keywords extraction, and importance weighting. Use when processing a job description to identify key requirements and categorize them for resume matching.

JD Analyze

You are tasked with analyzing a job description (JD) and extracting structured requirements from it.

Input

You receive the full text of a job description.

Output

Return a JSON object matching this schema:

{
  "title": "string — the job title",
  "company": "string | undefined — company name if mentioned",
  "location": "string | undefined — location if mentioned",
  "requirements": [
    {
      "id": "req_1",
      "text": "string — the requirement text",
      "category": "must_have | nice_to_have | culture | logistics",
      "importance": 0.0-1.0
    }
  ],
  "keywords": ["string — technical and domain keywords"],
  "experienceLevel": "string — e.g. 'Senior', 'Mid-level', 'Entry-level'"
}

Categorization Rules

Classify each requirement into exactly one category:

• must_have: Hard technical requirements, mandatory skills, required experience levels, required certifications. These are non-negotiable qualifications.
• nice_to_have: Preferred but not required skills, bonus qualifications, "plus" items. Often prefixed with "preferred", "nice to have", "bonus".
• culture: Team values, work style, communication expectations, cultural fit indicators. E.g. "collaborative environment", "fast-paced", "mentoring".
• logistics: Location, visa, clearance, travel, schedule, compensation information. E.g. "on-site in NYC", "US citizenship required".

ID Naming Convention

Assign sequential IDs: req_1, req_2, req_3, etc. Each distinct requirement gets its own ID — do not merge multiple requirements into one entry.

Importance Weighting

Assign importance scores (0.0 to 1.0) using these guidelines:

• 1.0: Explicitly stated as "required", "must have", or appears in a "Requirements" section header
• 0.8–0.9: Strongly implied requirements, repeated keywords, or in "Qualifications" section
• 0.5–0.7: Nice-to-have items, preferred qualifications
• 0.3–0.4: Cultural fit items, soft skills
• 0.1–0.2: Logistics items, unless they are deal-breakers (e.g. security clearance → 0.8)

Keywords Extraction

Extract 10–20 technical and domain-specific keywords that appear in the JD. Include:

• Programming languages and frameworks
• Tools and platforms
• Domain-specific terms (e.g. "distributed systems", "ML pipeline")
• Methodologies (e.g. "Agile", "CI/CD")

Do NOT include generic words like "team", "experience", "skills".

Edge Cases

• Vague JDs: If the JD is very short or vague, extract what you can and set importance conservatively.
• Multiple roles: If the JD describes multiple positions, focus on the primary role indicated by the title.
• Non-English JDs: Handle both English and Chinese job descriptions.

For detailed category definitions and calibration examples, consult references/TAXONOMY.md.