This document curates relevant skills and subagents from awesome-agent-skills and awesome-claude-code-subagents, tailored to the nomadically.work project with emphasis on remote EU job board aggregation and matching.
Purpose: Build responsive, performant job board interfaces with excellent UX for remote workers across EU markets.
- Frontend Design and UI/UX Development — Design-to-code workflows for job listings, filters, and application tracking
- Web Artifacts Builder with React/Tailwind — Rapid prototyping of job search, resume, and settings pages
- Next.js Optimization and Caching Strategies (Vercel) — Static generation for job listings, ISR for real-time updates
- nextjs-developer (voltagent-core-dev) — Framework expertise for App Router, API routes, middleware
- frontend-developer (voltagent-core-dev) — React 19 component architecture, state management
- react-specialist (voltagent-lang) — Advanced patterns for job filtering, search bars, infinite scroll
- ui-designer (voltagent-core-dev) — Design system consistency for multi-language job board (EU localization)
Purpose: Design scalable GraphQL API with D1 database for job aggregation, filtering, and resume matching across EU platforms.
- MCP Server Creation — Build Model Context Protocol servers for ATS integrations (Greenhouse, Lever, Ashby)
- Cloudflare Tools (Workers, D1, Vectorize, Queues) (Cloudflare Team) — Serverless job processing pipeline, resume RAG indexing, background job classification
- graphql-architect (voltagent-core-dev) — Schema design for jobs, companies, applications, skills taxonomy
- backend-developer (voltagent-core-dev) — Apollo Server 5 resolvers, context with Clerk auth, federation patterns
- database-administrator (voltagent-infra) — Drizzle ORM optimization, D1 query performance, migration strategy
- api-designer (voltagent-core-dev) — REST endpoint design for text-to-SQL, company enrichment, bulk imports
- database-optimizer (voltagent-data-ai) — Index optimization for job search queries, skill extraction performance
Purpose: Develop ML pipeline for accurate remote EU job classification, skill extraction, and resume matching.
- Prompt Engineering — Optimize prompts for job classification, bias detection, skill extraction accuracy
- Document Handling (PDF, DOCX) — Parse resumes and job descriptions for skill matching
- llm-architect (voltagent-data-ai) — Design multi-LLM pipelines (Claude, DeepSeek, Google) for job classification
- prompt-engineer (voltagent-data-ai) — Refine classification and extraction prompts; manage Langfuse prompt versions
- nlp-engineer (voltagent-data-ai) — Skill extraction, entity recognition for remote work indicators (flexible hours, timezone-agnostic, etc.)
- machine-learning-engineer (voltagent-data-ai) — Train classifiers for EU remote job signals; handle regional variations
- ai-engineer (voltagent-data-ai) — End-to-end ML pipeline management, model evaluation with Promptfoo/Vitest
Purpose: Ingest jobs from Greenhouse, Lever, Ashby; extract skills; maintain taxonomy; evaluate model accuracy.
- data-engineer (voltagent-data-ai) — Build ATS fetcher pipelines (Greenhouse, Lever, Ashby), ETL workflows
- data-analyst (voltagent-data-ai) — Job board coverage analysis, skill taxonomy validation, remote work trend reporting
- data-scientist (voltagent-data-ai) — Statistical analysis of job classifications, bias audit (fairness across EU regions)
Purpose: Deploy and maintain the application on Vercel + Cloudflare Workers with resilient data pipelines.
- Cloudflare Workers Deployment — D1 Gateway, ashby-crawler (Rust/WASM), process-jobs (Python/LangGraph)
- Web Performance Auditing (Cloudflare Team) — Job board page speed, Core Web Vitals monitoring
- devops-engineer (voltagent-infra) — CI/CD for Next.js + Workers, deployment orchestration
- deployment-engineer (voltagent-infra) — Vercel + Cloudflare multi-environment setup, secrets management
- platform-engineer (voltagent-infra) — Observability (Langfuse/LangSmith), error tracking, log aggregation
- docker-expert (voltagent-infra) — Containerize Python Workers if moving to non-CF platforms
- sre-engineer (voltagent-infra) — Incident response, job classification failures, database connectivity issues
Purpose: Evaluate job classification accuracy, detect bias, assess resume matching quality.
- Smart Contract Vulnerability Scanning (Trail of Bits) — Adapt security scanning patterns for LLM prompt injection
- Property-Based Testing (Trail of Bits) — Generate test cases for edge cases in skill extraction
- qa-expert (voltagent-qa-sec) — Test remote work classification accuracy across job descriptions
- test-automator (voltagent-qa-sec) — Automated Vitest suites for evals, regression tests for classifiers
- performance-engineer (voltagent-qa-sec) — Profile job listing queries, skill extraction latency, resume RAG search speed
- accessibility-tester (voltagent-qa-sec) — WCAG compliance for job board (essential for EU accessibility regulations)
Purpose: Align product with remote worker needs across EU markets; manage features and strategy.
- product-manager (voltagent-biz) — Feature prioritization, user feedback integration, remote work market research
- seo-specialist (voltagent-domains) — Job board SEO, geo-targeted keywords for EU remote job search
- content-marketer (voltagent-biz) — Blog posts, guides for remote EU job seekers, skill trend reports
- business-analyst (voltagent-biz) — Market analysis, competitor monitoring (other EU job boards), revenue modeling
- technical-writer (voltagent-biz) — Documentation for job ingestion, resume upload, API endpoints
Purpose: Secure user data (resumes, preferences), ensure GDPR compliance for EU operations.
- security-engineer (voltagent-infra) — Audit GraphQL queries for injection, Clerk auth hardening, API key rotation
- security-auditor (voltagent-qa-sec) — GDPR data retention audits, PII detection in job descriptions and resumes
- compliance-auditor (voltagent-qa-sec) — EU employment law compliance (working hours in job postings, anti-discrimination)
- Remote work signals: "fully remote," "work from anywhere," "hybrid," "flexible location," "async-friendly"
- EU timezone compliance: Offices or hours supporting CET/UTC±1 timezones
- Work-life balance: References to max working hours (EU Directive 2003/88), flexible schedules
- Language diversity: Multi-language job postings (indicator of EU-wide hiring)
- Visa/sponsorship: EU visa sponsorship, relocation support (common in EU remote roles)
- Data residency: Job data stored within EU (GDPR-friendly)
- Eastern EU: Often more remote-friendly (lower cost of living, timezone flexibility)
- Western EU: Hybrid patterns more common (London, Berlin, Amsterdam tech hubs)
- Southern EU: Growing remote adoption (Portugal, Spain remote visa programs)
- frontend-developer + nextjs-developer: Build job listing UI with filters
- graphql-architect + backend-developer: Design GraphQL schema, resolvers
- database-administrator: Optimize D1 schema for job queries
- prompt-engineer: Refine job classification prompts for "remote EU" signal detection
- nlm-architect: Set up multi-LLM routing (Claude for classification, DeepSeek for cost, etc.)
- data-engineer: Ingest jobs from ATS platforms using Trigger.dev/Inngest
- nlp-engineer: Extract remote work signals (keywords, timezone mentions, salary bands)
- ai-engineer: Build evaluation suite with Promptfoo/Vitest for classifier accuracy
- performance-engineer: Profile skill extraction, resume RAG latency
- devops-engineer: CI/CD pipeline for Next.js + Workers
- platform-engineer: Observability, error tracking, alerts for job ingestion failures
- sre-engineer: Incident response playbooks
- product-manager: Remote work feature prioritization, user surveys
- seo-specialist: Optimize job board for "remote jobs in [EU city]" search
- content-marketer: Publish remote work trend reports, salary benchmarks
- Awesome Agent Skills: https://github.com/VoltAgent/awesome-agent-skills
- Awesome Claude Code Subagents: https://github.com/VoltAgent/awesome-claude-code-subagents
- nomadically.work Project:
/home/user/nomadically.work - CLAUDE.md: Architecture, tech stack, data flow documentation