Job Posting Analysis Skill

Purpose

Extract technology stack information from job postings and career pages, which often reveal internal tech stack details.

Operations

1. find_careers_page

Locate company's career/jobs page.

Search Strategies:

1. Common paths: /careers, /jobs, /work-with-us, /join-us
2. Subdomains: careers.{domain}, jobs.{domain}
3. Web search: site:{domain} careers OR jobs
4. Footer links on main site

Common Career Page URLs:

https://{domain}/careers https://{domain}/jobs https://careers.{domain} https://jobs.{domain} https://{domain}/about/careers https://{domain}/company/careers

2. detect_ats_platform

Identify Applicant Tracking System in use.

ATS Detection Patterns:

{ "Greenhouse": { "url_pattern": "boards.greenhouse.io", "indicates": ["Tech-forward startup", "Modern hiring"], "confidence": 95 }, "Lever": { "url_pattern": "jobs.lever.co", "indicates": ["Tech-forward startup", "Growth stage"], "confidence": 95 }, "Workday": { "url_pattern": ".wd5.myworkdayjobs.com|.wd3.myworkdayjobs.com", "indicates": ["Enterprise company", "Large org"], "confidence": 95 }, "Ashby": { "url_pattern": "jobs.ashbyhq.com", "indicates": ["Modern startup", "Tech-forward"], "confidence": 95 }, "iCIMS": { "url_pattern": "careers-.*\.icims\.com|icims.com", "indicates": ["Enterprise hiring"], "confidence": 95 }, "Taleo": { "url_pattern": "taleo.net", "indicates": ["Enterprise (Oracle)", "Large org"], "confidence": 95 }, "SmartRecruiters": { "url_pattern": "jobs.smartrecruiters.com", "indicates": ["Mid-market to Enterprise"], "confidence": 95 }, "BambooHR": { "url_pattern": ".bamboohr.com/jobs", "indicates": ["SMB company"], "confidence": 95 }, "Jobvite": { "url_pattern": "jobs.jobvite.com", "indicates": ["Mid-market hiring"], "confidence": 95 }, "Breezy HR": { "url_pattern": ".breezy.hr", "indicates": ["SMB startup"], "confidence": 95 } }

3. extract_tech_requirements

Parse job descriptions for technology mentions.

Extraction Patterns:

Experience with ([\w\s,/]+) Proficiency in ([\w\s,/]+) Knowledge of ([\w\s,/]+) Tech stack:? ([\w\s,/]+) Working knowledge of ([\w\s,/]+) Familiar with ([\w\s,/]+) Strong background in ([\w\s,/]+) Required:?\s*([\w\s,/]+) Nice to have:?\s*([\w\s,/]+) Technologies:?\s*([\w\s,/]+) Tools:?\s*([\w\s,/]+)

Technology Keyword Categories:

Languages:

JavaScript, TypeScript, Python, Java, Go, Rust, Ruby, PHP, C#, C++, Kotlin, Swift, Scala, Elixir, Clojure

Frontend Frameworks:

React, Vue, Angular, Svelte, Next.js, Nuxt, Gatsby, Redux, MobX, Zustand, React Query, Tailwind, Bootstrap

Backend Frameworks:

Node.js, Express, NestJS, Django, Flask, FastAPI, Rails, Spring, .NET, Laravel, Phoenix

Databases:

PostgreSQL, MySQL, MongoDB, Redis, Elasticsearch, DynamoDB, Cassandra, Neo4j, Snowflake, BigQuery

Cloud/Infrastructure:

AWS, GCP, Azure, Kubernetes, Docker, Terraform, Ansible, CloudFormation, Pulumi

Tools:

Git, GitHub, GitLab, Jenkins, CircleCI, GitHub Actions, Datadog, New Relic, Grafana, Prometheus, Sentry

4. calculate_tech_frequency

Weight technologies by mention frequency across postings.

Scoring:

def calculate_frequency_score(tech, postings): mentions = sum(1 for p in postings if tech in p.requirements) total_postings = len(postings)

frequency = mentions / total_postings

# Classify importance
if frequency >= 0.5:
    importance = "Core Stack"  # 50%+ of postings
elif frequency >= 0.25:
    importance = "Common"      # 25-50%
else:
    importance = "Occasional"  # < 25%

return {
    "mentions": mentions,
    "frequency": frequency,
    "importance": importance
}

5. analyze_role_patterns

Identify tech stack from role types.

Role Type Signals:

{ "Frontend Engineer": { "implies": ["React/Vue/Angular", "JavaScript/TypeScript", "CSS frameworks"], "confidence": 70 }, "Backend Engineer": { "implies": ["Server-side language", "Database", "API development"], "confidence": 70 }, "Full Stack Engineer": { "implies": ["Frontend framework", "Backend framework", "Database"], "confidence": 65 }, "DevOps Engineer": { "implies": ["Cloud platform", "CI/CD", "Kubernetes/Docker", "IaC"], "confidence": 75 }, "Data Engineer": { "implies": ["Python/Scala", "Spark/Airflow", "Data warehouse"], "confidence": 75 }, "ML Engineer": { "implies": ["Python", "TensorFlow/PyTorch", "Cloud ML services"], "confidence": 75 }, "iOS Developer": { "implies": ["Swift", "Xcode", "iOS SDK"], "confidence": 85 }, "Android Developer": { "implies": ["Kotlin/Java", "Android SDK"], "confidence": 85 } }

Output

{ "skill": "job_posting_analysis", "domain": "string", "results": { "careers_page": { "url": "string", "ats_platform": "Greenhouse", "ats_confidence": 95 }, "postings_analyzed": "number", "technologies_extracted": [ { "name": "React", "category": "Frontend Framework", "mentions": 15, "total_postings": 20, "frequency": 0.75, "importance": "Core Stack", "contexts": [ "Experience with React and TypeScript", "Build UIs using React" ], "confidence": 80 } ], "role_distribution": { "Frontend": 5, "Backend": 8, "Full Stack": 4, "DevOps": 2, "Data": 1 }, "tech_stack_inference": { "frontend": ["React", "TypeScript", "Tailwind"], "backend": ["Node.js", "PostgreSQL", "Redis"], "infrastructure": ["AWS", "Kubernetes"], "confidence": "Medium" }, "company_signals": { "engineering_size": "Large (20+ open roles)", "growth_stage": "Scaling", "tech_culture": "Modern (tech-forward ATS, current stack)" } }, "evidence": [ { "type": "job_posting", "title": "Senior Frontend Engineer", "url": "string", "technologies_mentioned": ["React", "TypeScript", "GraphQL"], "timestamp": "ISO-8601" } ] }

Rate Limiting

• Careers page fetch: 10/minute

• Job posting pages: 20/minute

• ATS APIs: Varies by platform

Error Handling

• 404: No careers page found

• Access denied: ATS may require authentication

• Continue with partial data

• Fall back to search engine results

Security Considerations

• Only access public job postings

• Do not apply to jobs or create accounts

• Respect robots.txt

• Do not scrape PII (recruiter names, emails)

• Log all fetches for audit

Confidence Notes

Job posting data provides indirect signals:

• Technologies mentioned in job posts may not be currently deployed

• "Nice to have" vs "Required" distinction matters

• Combine with direct technical evidence for validation

• Base confidence: 60-80% (lower than direct signals)