werlang

Build or refactor reusable Node.js APIs using Express, express.Router, ESM imports, named exports, entity classes in model/, helpers for non-endpoint logic, middleware for auth and other pre-route concerns, and route handlers that talk to model instances. Use when creating CRUD APIs, JWT auth flows, bcrypt password login, route modules, or starter API structures.

Deliver behavior changes, bug fixes, refactors, and feature work in any project with tests or explicit validation. Use when changing application behavior, updating existing tests, writing tests, reviewing missing coverage, choosing between automated and manual validation, or documenting testing gaps across backend, frontend, CLI, and database migrations.

Build or refactor reusable server-rendered web frontends using Express, render middleware, template variables, page entry files in src/js root, DOM-related component classes in components/, support helpers in helpers/, and API-backed entity classes in models/. Use when scaffolding or reorganizing frontend projects that consume an API through frontend model methods.

Codebase Search & Intelligence using SocratiCode. Use when exploring code structure, searching concepts, finding symbol usages, inspecting impact/blast radius before refactoring, tracing execution flows, checking circular dependencies, or searching non-code context artifacts. Automatically checks index status and triggers indexing if the project is not yet indexed.

Enforce reusable, brand-aware CSS architecture for public web apps by owning shared tokens in a common stylesheet, importing page entrypoints first, keeping component partials scoped, and using consistent media-query and color-token conventions.

Maintain and query the developer's persistent Obsidian vault memory, component registry, architecture decision records (ADRs), and dev changelog graph. Automatically invoke whenever creating or modifying relevant code (features, fixes, refactors, shared components), making architectural decisions, capturing technical brainstorming, or when the user references past work or component versions across projects.

Use when the user wants to design, redesign, shape, critique, audit, polish, clarify, distill, harden, optimize, adapt, animate, colorize, extract, or otherwise improve a frontend interface. Covers websites, landing pages, dashboards, product UI, app shells, components, forms, settings, onboarding, and empty states. Handles UX review, visual hierarchy, information architecture, cognitive load, accessibility, performance, responsive behavior, theming, anti-patterns, typography, fonts, spacing, layout, alignment, color, motion, micro-interactions, UX copy, error states, edge cases, i18n, and reusable design systems or tokens. Also use for bland designs that need to become bolder or more delightful, loud designs that should become quieter, live browser iteration on UI elements, or ambitious visual effects that should feel technically extraordinary. Not for backend-only or non-UI tasks.

Deeply review frontend code for rendering regressions, interaction bugs, state and URL sync defects, auth and redirect issues, accessibility gaps, date/time problems, bundle or template contract drift, and missing regression coverage. Use when auditing UI code, investigating user-facing bugs, or pairing findings with deterministic frontend tests and honest browser validation.

Enforce the current web CSS architecture and warm editorial styling conventions in this repository. Use when creating or changing styles, theming, design tokens, component visuals, layout classes, or page atmosphere in web/src/css.

Build, review, and maintain the Academic Events web app (Express SSR shell plus bundled vanilla JavaScript UI). Use when changing or reviewing the home, login, week, or dashboard pages, client auth and event flows, frontend tests, browser validation, styles, webpack behavior, static serving, or API integration from the web layer.

Build and maintain REST API endpoints for authentication, moderation, notifications, and academic events in `api/`. Use when adding or modifying routes, request validation, auth flows, response contracts, background side effects, model behavior, or MySQL-backed persistence.

Add focused JSDoc and comments to touched reusable Event Hub code while avoiding noisy narration of obvious statements.

Keep Event Hub documentation, agent guidance, prompts, skills, comments, and validation instructions synchronized with the real implementation after code, route, config, CSS, build, Compose, or test changes.

Deliver feature work, frontend test design, and refactors in this Academic Events repository with tests or explicit validation. Use when changing behavior, adding features, fixing regressions, updating tests, reviewing missing coverage, validating browser interactions, or documenting testing gaps across api/ and web/.

Deeply review the Academic Events API in `api/` for logic bugs, contract mismatches, authorization gaps, state-transition defects, background-task failures, time/date bugs, and missing regression coverage. Use when auditing API code for bugs or pairing bug-finding with deterministic Jest test creation and reruns.

Diagnose runtime and integration issues across API and Web services in this Academic Events project. Use when services fail to boot, auth or event flows break, API responses are unexpected, frontend rendering is incorrect, or Docker development behavior is unstable.

Build and maintain the API and Worker services for this LLM streaming relay project. Use when adding or modifying routes, request validation, SSE relay logic, queue behavior, WebSocket server events, LLM model runner integration, or marketplace order/settlement flows.

Write and run tests for the API and Worker services using Node.js built-in node:test. Use when adding new tests, extending coverage, fixing failing tests, choosing between unit and integration test strategies, or reviewing test gaps identified in audits.

Diagnose runtime and integration issues across the API and Worker services. Use when services fail to boot, workers do not connect, LLM jobs stall, SSE streams are silent, database operations fail, or Docker networking breaks.

Audit, verify, update, or clean docs, prompts, skills, agents, instructions, README files, and other markdown/MD files against the current project codebase. Use when documentation or agent context may be stale, inaccurate, inconsistent, or aspirational and must be aligned with the real implementation.

Deeply review api/ and worker/ service code for WebSocket session bugs, SSE relay defects, queue dispatch errors, job lifecycle bugs, and LLM stream processing failures. Use when auditing server-side code for real defects or pairing bug-finding with deterministic regression tests.

Configure and run the API and Worker services with Docker Compose. Use when starting, stopping, rebuilding, or troubleshooting either service container, or when investigating networking, port, environment variable, or WebSocket connectivity issues.

Update README.md, .github/memories/, and .github/skills/ after any new feature, refactor, or code behavior change in api/ or worker/. Auto-invoke when code changes are finalized. Trigger phrases: "update docs", "update memories", "keep docs updated", "docs are stale", "update skills".

Document touched code with JSDoc, method comments, and focused section comments as the default behavior whenever implementing a new feature, refactoring, changing a body of code, writing tests, or improving maintainability. Apply this by default for all code-writing tasks; if the user explicitly requests no comments, respect that.

Build or change API behavior in api/ using Express routers, middleware, CustomError, model instances, and camelCase JSON responses.

Review backend and frontend template code for real defects, boundary violations, security gaps, and missing tests. Auto-invoke for code reviews or bug investigations. Triggers: "bug review", "review code", "find bugs", "security audit", "check defects".

Implement OOP API entities in api/model/ using the shared Model base class and relation helper.

Maintain the api/helpers/postgres.js SQL boundary, CRUD helper surface, and safe filter conventions.

Update project-local agent guidance when architecture, workflows, commands, or conventions change. Auto-invoke when user or agent output indicates a reusable workflow or durable rule. Triggers: "update docs", "update skill", "create new skill", "update agent instructions", "document recurring pattern", "capture reusable workflow".

Add accurate JSDoc and focused comments when changing template code. Auto-invoke during code writing or refactoring. Triggers: "document code", "jsdoc", "code comments", "add comments", "explain logic".

Keep human docs, agent guidance, and touched code comments aligned with the real web/api implementation, compose workflow, and validation commands. Auto-invoke when auditing docs or after major architectural changes. Triggers: "audit docs", "check context", "documentation maintenance", "audit documentation", "verify docs".

Deliver template behavior changes with focused automated tests, build checks, and honest validation reporting. Auto-invoke when writing tests, fixing bugs, refactoring routes/models, or validating web/api changes. Triggers: "test-first", "run test", "add unit test", "add integration test", "browser smoke", "validate changes".

Parecer academico no papel de avaliador, banca ou peer reviewer. Use para avaliar TCCs, artigos e capitulos sem editar o texto por padrao, classificando problemas por severidade, evidencias, criterios, nota/rubrica quando solicitada e recomendacoes acionaveis.

Revisao e escrita de sintaxe LaTeX academica, ABNT e template de TCC. Use para corrigir figuras, tabelas, captions, labels, refs, BibTeX, comandos do template, pre-textuais, apendices, citacoes e problemas de compilacao em projetos LaTeX academicos.

Busca e triagem de fontes academicas para TCCs e artigos. Use para encontrar trabalhos relacionados, fontes recentes, referencias primarias, PDFs, DOI, metadados, lacunas de pesquisa e candidatos a citacao antes da verificacao com bibtex-verified-citations.

Gerenciamento de BibTeX com verificacao obrigatoria de citacoes na internet. Use para inserir, auditar ou corrigir referencias em .bib e no texto, garantindo que a fonte consultada sustente a afirmacao citada e registrando URL/evidencia em project/citation-cheatsheet.md.

Workflow especifico do workspace Overleaf Hero para revisar, orientar ou avaliar TCCs em LaTeX. Use quando trabalhar no diretorio project/, mapear tcc.tex, preservar template/sample*, aplicar prompts de autor/orientador/avaliador e coordenar as skills academicas do projeto.

Orientacao de pesquisa cientifica para TCCs, artigos e dissertacoes. Use quando o usuario pedir feedback de orientador, ajuda com problema de pesquisa, objetivos, lacuna, metodologia, resultados, limitacoes, proximos passos ou planejamento de escrita.

Proofreading cientifico em PT-BR para TCCs, artigos e dissertacoes. Use para corrigir ortografia, gramatica, pontuacao, clareza, coesao, impessoalidade, terminologia e fluidez preservando a voz do autor e sem alterar conteudo tecnico.

Escrita e reescrita de texto cientifico em LaTeX. Use para construir secoes academicas, transformar notas em parágrafos, melhorar argumentacao, estruturar introducao/fundamentacao/metodologia/resultados/conclusao, ou escrever corpo de texto preservando evidencias e voz do autor.

Build and update TMET backend APIs in Express. Use when changing handlers in routes/*.js, file-backed business logic in models/*.js, queue-driven async jobs, scraping/upload integrations in helpers/*.js, or response contracts under /api/matches, /api/moodle, /api/suap, /api/ai, and /api/jobs.

Validate TMET API contracts and multi-step flows, especially async job polling, generated file artifacts, and extraction -> matching -> enrollment workflows. Use when verifying behavior across routes, models, helpers, queue progress, and persisted outputs in files/.

Write and maintain TMET Jest + Supertest tests. Use when changing routes, models, helpers, queue logic, file parsing, or API contracts and you need focused coverage under tests/routes, tests/models, tests/helpers, or tests/integration.

Diagnose TMET runtime, queue, scraping, upload, and file-artifact issues. Use when investigating failing routes, stuck jobs, broken Browserless/SUAP flows, malformed CSV or JSON outputs, Docker problems, or test regressions.

Run, debug, and validate TMET with Docker Compose. Use when bringing up the current node + chrome stack, checking logs, running tests in-container, or confirming Browserless-based scraping and upload workflows.

Operate and evolve the TMET extraction to matching to CSV to Moodle-upload pipeline. Use when changing or debugging EduPage extraction, SUAP scraping, manual course or student queues, subject matching, AI matching, student/professor extraction, generated files in files/, or end-to-end operator workflow behavior.

Implement and maintain the TMET static frontend in public/. Use when changing public/index.html, browser-native ES modules in public/js/, UI components, section controllers, or request/model adapters that drive the pipeline, matching, manual queue, and students/professors workflows.

Use esta skill ao criar ou revisar exercícios práticos deste repositório, especialmente em seções didáticas de HTML, CSS, JavaScript, DOM, arrays, objetos e PHP. Ela ajuda a produzir enunciados concretos em PT-BR, com objetivos claros, requisitos operacionais, reaproveitamento de conteúdos anteriores e estrutura compatível com a organização pedagógica do projeto.

Create, revise, refactor, or explain source code for this repository in a didactic, beginner-friendly style for high-school students. Use when Agent needs to write or change HTML, CSS, JavaScript, PHP, or code snippets in lessons, examples, exercises, mini-projects, or avaliações, keeping the implementation educational, explicit, well documented, and with comments in PT-BR.

Transform a didactic README guide into a concise Marp slide presentation for class, aligned with the ifsul theme and the pattern of marp/content/00-introducao.md. Use when Agent needs to create, revise, or expand lecture slides from a README, summarize a long guide for aula, montar apresentacao Marp, criar slide de apoio para o professor, or produce a short teaching presentation with image placeholders.

Transform section, lesson, topic, or mini-project README files into clear didactic reference guides in Portuguese, with consistent structure, small examples, and student-friendly explanations. Use when Agent needs to create, expand, rewrite, or standardize a README for aula, material de apoio, referencia conceitual, guia pratico, enunciado de exercicio, documentacao de API, or when requests mention: melhorar README, transformar slide em guia, criar guia de JS/DOM/eventos/PHP/fetch, padronizar documentacao didatica.

Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Agent's capabilities with specialized knowledge, workflows, or tool integrations.

Use when changing AutoJudge Contest VS Code extension runtime behavior, including login/logout flows, session restore, contest tree rendering, problem and submission actions, submission workflows, output-channel rendering, or AutoJudge API error handling.

Use when changing the AutoJudge Contest VS Code extension manifest surface, including views, commands, menus, settings, context keys, or user-facing documentation that must stay synchronized with package.json.

Use when a task needs repository commands or local tooling so the work runs through the Compose development container instead of the host machine.

Use when refactoring or changing behavior in this repository so work follows a Vitest-first workflow executed inside the Compose development container.

Human-like long-term memory across sessions. Run memory_search at task start unless the user opts out, run additional searches only on mid-task triggers, persist reusable outcomes with memory_write, and prune stale or duplicate entries with memory_forget so active memory stays small and relevant.

Enforce the current web CSS architecture and warm editorial styling conventions in this repository. Use when creating or changing styles, theming, design tokens, component visuals, layout classes, or page atmosphere in web/src/css.

Add accurate JSDoc and focused comments when changing portfolio code.

Build or change the lean Express/Mustache/Webpack portfolio in web/, including the static build path, page script, and checked-in public assets.

Update project-local agent guidance when architecture, workflows, commands, or conventions change.

Add, refactor, or debug backend WebSocket routes, handlers, router, queue logic, and report PDF generation. Use when implementing new WS routes in api/app.js or updating the document generation helper in api/helpers/document-builder.js.

Implement, debug, or style web frontend wizard steps, UI components, state management, and LESS stylesheets. Use when modifying the 3-step wizard flow, editing UI cards, updating toggles, or editing LESS files under mobile-first and variable-only CSS rules.

Debug, maintain, and update the Playwright-based SUAP scraping system in api/helpers/scraper.js and configuration selectors in api/suap-config.js. Use when investigating scraping errors, fixing login hangs, adjusting selector/navigation timeouts, or parsing new academic data fields from SUAP pages.

Use when adding or modifying AutoJudge API routes, model classes, auth checks, MySQL-backed entity behavior, or API Jest coverage. Covers the route-model-helper-test workflow used in api/.

Use when validating AutoJudge API, web, judge, background, or Docker Compose changes. Covers the narrowest honest command selection for Jest, Webpack rebuilds, and container-aware checks.

Use when modifying AutoJudge web pages, Mustache views, template descriptors, browser component classes, frontend models, or Webpack-built assets in web/.

Use when changing the VS Code extension manifest surface for AutoJudge, including commands, keybindings, menus, settings, supported languages, or user-facing documentation that must stay synchronized with package.json.

Use when changing AutoJudge VS Code extension runtime behavior, including the run command flow, input sidecar resolution, API queue and poll logic, output channel rendering, cancellation, or AutoJudge error handling.