Skillenai API Skill
skillenai/skillenai-api-skill
Inteligência de mercado de trabalho AI/ML para busca de vagas e análise de skills
Ver no GitHubSkill para Claude que integra a API de inteligência de mercado de trabalho Skillenai, fornecendo buscas de empregos, análise de skills, tendências de tópicos e resolução de entidades no ecossistema AI/ML. Permite agentes acessarem dados estruturados sobre vagas, requisitos técnicos, salários e relações entre jobs, skills e empresas.
Instalação
npx convez-ai skillenai/skillenai-api-skill1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.
Métricas
Casos de uso
- Buscar vagas alinhadas a um perfil de resume específico
- Analisar quais skills são demandados para um determinado cargo
- Acompanhar tendências de tecnologias e tópicos no mercado AI/ML
- Criar planos de aprendizado personalizados com base em análise de mercado
Veredito

Skill bem construída e documentada que oferece acesso estruturado a dados reais de mercado de trabalho. Ideal para agentes que precisam fazer recomendações de carreira, análise de skills ou pesquisa de mercado no ecosistema AI/ML, com autenticação segura via OAuth.
README
A tradução para o português deste README ainda está na fila.
Skillenai API Skill
A Claude Code skill and Python toolkit for the Skillenai Data Products API — labor market intelligence covering skills, jobs, trends, and entity analytics across the AI/ML ecosystem.
What Skillenai Provides
Skillenai indexes and enriches content from across the AI/ML landscape — job postings, blog posts, research papers, news, and social media — then extracts entities, resolves them to a canonical knowledge graph, and exposes the results through a structured API.
Key Capabilities
- Job Search — Multi-signal ranking combining text relevance, skill matching, seniority, salary, location, and recency
- Skill Analysis — What skills does a given role require? Compare skill profiles across roles with fuzzy role name resolution
- Topic Trends — Track how topics rise and fall in the AI/ML discourse over time
- Entity Resolution — Map free-text names ("PyTorch", "pytorch", "Py Torch") to canonical entity IDs
- Knowledge Graph — Traverse relationships between jobs, skills, companies, people, and products via Cypher queries
Quick Start
- Register at app.skillenai.com and verify your email
- Create an API key in the dashboard — it will look like
skn_live_... - Make your first request:
curl -s -H "X-API-Key: skn_live_your_key_here" \
"https://api.skillenai.com/v1/analytics/counts" | python3 -m json.tool
See docs/getting-started.md for the full walkthrough.
If you're using the Claude Code plugin, you don't need to copy a key by hand — run /skillenai:api setup and authorize in the browser. The Credentials section below explains both paths.
Use with Claude Code
This repo is packaged as a Claude Code plugin. The recommended install is through the plugin marketplace:
/plugin marketplace add skillenai/skillenai-api-skill
/plugin install skillenai@skillenai-api
Run these as two separate commands — Claude Code slash commands are one per line. Restart Claude Code (or /reload-plugins) after installing. The plugin registers three skills:
/skillenai:api— query the data products API (skills, jobs, trends, entity analytics)/skillenai:job-finder— match a resume against the live job index/skillenai:blog— author and manage Skillenai blog posts from inside Claude
Note the colon — standard namespacing for plugin skills.
Credentials
The recommended path is the in-skill OAuth flow:
/skillenai:api setup
This opens app.skillenai.com/activate in your browser, you sign in (or sign up) and click Allow, and the issued key is written to ~/.skillenai/.env with mode 0600. The key is never printed to the terminal or to the conversation transcript — the agent only sees a ✓ Authorized confirmation.
Behind the scenes the skill calls the API through scripts/api.py, which loads ~/.skillenai/.env in its own process and signs requests with X-API-Key. The agent's shell never sees the key, so it can't be accidentally echoed, logged, or shown in ps.
If you'd rather set it up by hand (e.g. you already have a key from the dashboard, or you're scripting against the API outside the plugin):
mkdir -p ~/.skillenai
printf 'API_KEY=skn_live_your_key_here\n' > ~/.skillenai/.env
chmod 600 ~/.skillenai/.env
Or export it in your shell profile: export API_KEY=skn_live_.... Credentials resolve with the same precedence in every entry point: shell env → ~/.skillenai/.env → plugin-local .env → cwd .env.
Get an API key by registering at app.skillenai.com.
Asking questions
Once installed, try things like:
- "What skills do Data Scientists need?"
- "Search for remote ML engineer jobs paying over $200k"
- "Show me topic trends for the last 6 months"
- "Run a full EDA report"
Alternative: clone directly as a directory skill
If you prefer not to use the plugin system, you can register the skill manually by cloning the inner skills/api/ directory into a skills folder Claude Code scans:
git clone https://github.com/skillenai/skillenai-api-skill.git
ln -s "$(pwd)/skillenai-api-skill/skills/api" ~/.claude/skills/skillenai-api
Invocation in that layout is /skillenai-api (bare — directory skills aren't namespaced). The same ~/.skillenai/.env credentials work.
Alternative: use the repo standalone
Everything in this repo — skills/api/SKILL.md, scripts/, and docs/ — is just files. Clone it anywhere and point your own agent at the SKILL.md or run the Python scripts directly with python scripts/eda_report.py etc.
Helper Scripts
Python scripts in scripts/ provide CLI access to common workflows. Install dependencies:
pip install requests python-dotenv
| Script | Purpose |
|---|---|
scripts/oauth_setup.py | Browser-based OAuth setup that writes ~/.skillenai/.env |
scripts/api.py | Authenticated request wrapper used by the skill (keeps the key out of the agent's shell) |
scripts/eda_report.py | Generate a comprehensive EDA markdown report |
scripts/skill_analysis.py | Analyze skill demand by role, compare roles |
scripts/trend_analysis.py | Topic trend time series and growth analysis |
scripts/job_search.py | Multi-signal job search with formatted output |
scripts/blog_upload.py | Upload a cover image for a blog post (presign + PUT) |
scripts/scholarly_blog_phrase_ratio.py | Per-skill academic-depth ratio (scholarly / scholarly+blog match_phrase counts) — 9s-throttled, resumable |
scripts/phrase_prevalence.py | Document-prevalence of a long concept list across one or more indices, batched via a filters aggregation (one request per ~10 concepts); supports multi-spelling OR-groups |
Run any script with --help for usage.
Documentation
- Getting Started — Register, get a key, make your first request
- Authentication — API key format, headers, .env setup
- Rate Limits — Tiers, credits, billing headers, error codes
- Endpoints:
- Analytics — Counts, topic trends, entity co-occurrence, skills by role
- Jobs — Multi-signal job search
- Resolution — Entity resolution (free-text to canonical)
- Catalog — Schema introspection
- Query — SQL, OpenSearch DSL, Cypher graph queries
- Content — Blog post authoring (used by the
/skillenai:blogskill)