job-analysis

name: job-analysis description: | Parses job postings to extract requirements, signals, and red flags. Use when processing new job listings to understand what the role actually requires.

Job Analysis Skill

Overview

This skill parses job postings to extract structured data about requirements, signals, and red flags. It transforms unstructured job descriptions into actionable intelligence for downstream skills (profile-matching, fit-scoring).

When to Use

• When a new job posting enters the system
• When a user wants to understand what a job actually requires
• Before running profile-matching or fit-scoring
• When comparing multiple job postings

Workflow

Step 1: Extract Basic Information

Parse the job posting for:

Step 2: Extract Requirements

Categorize requirements into three tiers:

Must-Have Requirements

Skills/experience explicitly marked as "required":

• Years of experience
• Specific technologies
• Degree requirements
• Certifications

Preferred Requirements

Skills/experience marked as "preferred", "nice-to-have", or "bonus":

• Additional technologies
• Industry experience
• Soft skills

Implied Requirements

Requirements inferred from context:

• Team size from "lead a team of X"
• Seniority from responsibilities
• Culture fit from language used

Step 3: Identify Signals

Positive Signals

• Remote-first or distributed team mentions
• Equity/stock options mentioned
• Growth opportunities
• Modern tech stack
• Work-life balance mentions
• Parental leave highlighted
• Engineering culture emphasis

Negative Signals (Red Flags)

• "Fast-paced environment" (may indicate burnout culture)
• "Wear many hats" at large company (role confusion)
• Unclear responsibilities
• No salary range disclosed
• Excessive requirements (kitchen sink job posting)
• "Unlimited PTO" (may mean no PTO)
• "We're like a family" (boundary issues)

Step 4: Keyword Analysis

Extract and prioritize keywords by frequency:

Primary Keywords (3+ mentions):
- React, TypeScript, Node.js

Secondary Keywords (2 mentions):
- AWS, PostgreSQL, GraphQL

Tertiary Keywords (1 mention):
- Docker, Redis, CI/CD

Step 5: Role Classification

Classify the role type:

Output Format

{
  "job_id": "string",
  "parsed_at": "ISO8601",
  "basic_info": {
    "title": "string",
    "company": "string",
    "location": "string",
    "remote_policy": "fully_remote | hybrid | onsite",
    "salary_range": {
      "min": "number | null",
      "max": "number | null",
      "currency": "string"
    },
    "employment_type": "full_time | part_time | contract",
    "experience_level": "junior | mid | senior | staff | principal"
  },
  "requirements": {
    "must_have": [
      {
        "skill": "string",
        "years": "number | null",
        "category": "technical | domain | soft"
      }
    ],
    "preferred": [],
    "implied": []
  },
  "keywords": {
    "primary": ["string"],
    "secondary": ["string"],
    "tertiary": ["string"]
  },
  "signals": {
    "positive": [
      {
        "signal": "string",
        "evidence": "string"
      }
    ],
    "negative": [
      {
        "signal": "string",
        "severity": "warning | red_flag",
        "evidence": "string"
      }
    ]
  },
  "role_classification": {
    "type": "ic | tech_lead | manager | founding | staff",
    "confidence": "number (0-1)"
  },
  "analysis_notes": "string"
}

Integration Points

• Input: Raw job posting text from aggregator
• Output: Structured job analysis JSON
• Downstream: profile-matching, fit-scoring, company-evaluation

Quality Checklist

Before outputting analysis:

• All basic info fields extracted or marked null
• Requirements categorized into 3 tiers
• At least 3 keywords identified
• Both positive and negative signals evaluated
• Role type classified with confidence score
• Analysis notes explain any ambiguities