opencode-skills-collection 3.0.2 → 3.0.4

package/bundled-skills/andrej-karpathy/SKILL.md

@@ -108,7 +108,7 @@ A combinação que Karpathy representa é genuinamente rara:
 
 ---
 
-## 2.1 — Software 2.0
+### 2.1 — Software 2.0
 
 Publicado no Medium em 2017, este é o ensaio mais original e influente de Karpathy.
 A tese central mudou como a comunidade pensa sobre o que é programação:
@@ -147,7 +147,7 @@ the training data and curate the dataset. We are reprogramming computers with da
 **Com LLMs (2023):** Dataset = internet inteira. Loss = cross-entropy no próximo token.
 Emergência de capacidades que ninguém especificou explicitamente. Software 2.0 em escala máxima.
 
-## 2.2 — Llms Como Sistema Operacional
+### 2.2 — Llms Como Sistema Operacional
 
 Esta analogia, desenvolvida em 2023 (especialmente na palestra "State of GPT" no
 Microsoft Build), reframeu como pensar em LLMs como plataforma:
@@ -182,7 +182,7 @@ entendem linguagem natural e podem executar tarefas complexas quando instruídos
 em inglês, então inglês se tornou literalmente uma linguagem de programação —
 uma que qualquer falante nativo já "sabe", sem precisar aprender sintaxe especial.
 
-## 2.3 — Bottom-Up Learning (Filosofia Pedagógica Central)
+### 2.3 — Bottom-Up Learning (Filosofia Pedagógica Central)
 
 A regra mais importante: construa do zero antes de usar a biblioteca. Entenda a
 abstração antes de depender dela.
@@ -205,7 +205,7 @@ o estudante entende cada componente de qualquer LLM moderno.
 **Citação:** "The library is just convenience; the math is the substance. Once you
 understand how backprop works, you can use PyTorch with full confidence."
 
-## 2.4 — Vibe Coding
+### 2.4 — Vibe Coding
 
 Termo cunhado por Karpathy em fevereiro de 2025 em um tweet que viralizou na
 comunidade de programação. Define uma nova modalidade de desenvolvimento de
@@ -243,7 +243,7 @@ pequenos e exploratórios: superpotência. Para engenharia séria: ainda precisa
 pessoas que entendem o código. Mesmo "vibers" se beneficiam de fundamentos sólidos —
 para reconhecer quando o LLM gerou algo incorreto.
 
-## 2.5 — Scaling Laws E Emergência
+### 2.5 — Scaling Laws E Emergência
 
 **O que são scaling laws:** Relações empíricas mostrando que performance melhora
 previsível e regularmente com mais parâmetros (N), mais dados (D), mais compute (C).
@@ -264,7 +264,7 @@ teoricamente melhor que não escala em hardware disponível.
 
 ---
 
-## 3.1 — Contexto E Missão
+### 3.1 — Contexto E Missão
 
 Karpathy entrou na Tesla em junho de 2017 como Director of AI, assumindo
 responsabilidade pela equipe de visão e machine learning do Autopilot.
@@ -276,7 +276,7 @@ faixa para uma arquitetura de visão end-to-end capaz de condução autônoma em
 condições gerais. A stack construída foi a mais complexa e sofisticada de visão
 computacional já deployada em escala de produção massiva.
 
-## 3.2 — A Decisão Cameras-Only (Vs Lidar)
+### 3.2 — A Decisão Cameras-Only (Vs Lidar)
 
 Este é talvez o debate técnico mais importante da carreira de Karpathy, e ele
 articulou o argumento com precisão cirúrgica:
@@ -314,7 +314,7 @@ articulou o argumento com precisão cirúrgica:
 - A abordagem cameras-only coloca peso enorme na rede neural — funciona se e
   somente se a rede for suficientemente boa, o que é uma aposta high-stakes.
 
-## 3.3 — Hydranet: Uma Rede Para Tudo
+### 3.3 — Hydranet: Uma Rede Para Tudo
 
 Apresentado no Tesla AI Day (agosto 2021), o HydraNet é a arquitetura central
 de visão da Tesla descrita por Karpathy:
@@ -351,7 +351,7 @@ Backbone ──────────┼─── Head: Velocity Estimation (v
    um espaço de features compartilhado — o modelo "vê" o mundo 360° como um único
    volume de features, não como imagens separadas.
 
-## 3.4 — A Data Engine: O Produto Real
+### 3.4 — A Data Engine: O Produto Real
 
 O conceito mais sofisticado que Karpathy desenvolveu e articulou na Tesla.
 Sua tese: o modelo de produção não é o produto. A data engine — o sistema de
@@ -392,7 +392,7 @@ loop fechado entre frota, anotação e treinamento — é o produto.
 
 **Citação:** "The data engi
 
-## 3.5 — Dojo: Supercomputador Para Visão
+### 3.5 — Dojo: Supercomputador Para Visão
 
 Anunciado no Tesla AI Day 2021, Dojo foi o supercomputador proprietário da Tesla
 para treinamento de modelos de visão. Karpathy foi central na visão técnica:
@@ -403,7 +403,7 @@ para treinamento de modelos de visão. Karpathy foi central na visão técnica:
 - A decisão de construir hardware próprio reflete a filosofia de controle da stack
   que tanto Karpathy quanto Musk defendem
 
-## 3.6 — O Que Karpathy Aprendeu Na Tesla
+### 3.6 — O Que Karpathy Aprendeu Na Tesla
 
 Em entrevistas e tweets após sair, Karpathy articulou as lições mais importantes:
 
@@ -424,7 +424,7 @@ Em entrevistas e tweets após sair, Karpathy articulou as lições mais importan
 
 ---
 
-## 4.1 — Micrograd
+### 4.1 — Micrograd
 
 **Repositório:** github.com/karpathy/micrograd
 **Tamanho:** ~100 linhas de Python puro

package/bundled-skills/angular-best-practices/SKILL.md

@@ -557,10 +557,6 @@ export class Component implements OnInit, OnDestroy {
 - [Zoneless Angular](https://angular.dev/guide/experimental/zoneless)
 - [Angular SSR Guide](https://angular.dev/guide/ssr)
 - [Change Detection Deep Dive](https://angular.dev/guide/change-detection)
-
-## When to Use
-This skill is applicable to execute the workflow or actions described in the overview.
-
 ## Limitations
 - Use this skill only when the task clearly matches the scope described above.
 - Do not treat the output as a substitute for environment-specific validation, testing, or expert review.

package/bundled-skills/apify-actorization/SKILL.md

@@ -115,7 +115,7 @@ See [schemas-and-output.md](references/schemas-and-output.md) for detailed confi
 
 Validate schemas against `@apify/json_schemas` npm package.
 
-## Step 7: Test Locally
+## Step 4: Test Locally
 
 Run the actor with inline input (for JS/TS and Python actors):
 
@@ -131,7 +131,7 @@ apify run --input-file ./test-input.json
 
 **Important:** Always use `apify run`, not `npm start` or `python main.py`. The CLI sets up the proper environment and storage.
 
-## Step 8: Deploy
+## Step 5: Deploy
 
 ```bash
 apify push

package/bundled-skills/audit-context-building/SKILL.md

@@ -37,7 +37,7 @@ Do **not** use for:
 
 ---
 
-## 3. How This Skill Behaves
+## 2. How This Skill Behaves
 
 When active, Claude will:
 - Default to **ultra-granular analysis** of each block and line.
@@ -64,7 +64,7 @@ Goal: **deep, accurate understanding**, not conclusions.
 
 ---
 
-## 4. Phase 1 — Initial Orientation (Bottom-Up Scan)
+## 3. Phase 1 — Initial Orientation (Bottom-Up Scan)
 
 Before deep analysis, Claude performs a minimal mapping:
 
@@ -78,7 +78,7 @@ This establishes anchors for detailed analysis.
 
 ---
 
-## 5. Phase 2 — Ultra-Granular Function Analysis (Default Mode)
+## 4. Phase 2 — Ultra-Granular Function Analysis (Default Mode)
 
 Every non-trivial function receives full micro analysis.
 
@@ -199,7 +199,7 @@ Analysis is complete when all checklist items are satisfied and no unresolved "u
 
 ---
 
-## 6. Phase 3 — Global System Understanding
+## 5. Phase 3 — Global System Understanding
 
 After sufficient micro-analysis:
 
@@ -227,7 +227,7 @@ These clusters help guide the vulnerability-hunting phase.
 
 ---
 
-## 7. Stability & Consistency Rules
+## 6. Stability & Consistency Rules
 *(Anti-Hallucination, Anti-Contradiction)*
 
 Claude must:
@@ -255,7 +255,7 @@ Claude must:
 
 ---
 
-## 8. Subagent Usage
+## 7. Subagent Usage
 
 Claude may spawn subagents for:
 - Dense or complex functions.
@@ -275,7 +275,7 @@ Subagents must:
 
 ---
 
-## 9. Relationship to Other Phases
+## 8. Relationship to Other Phases
 
 This skill runs **before**:
 - Vulnerability discovery
@@ -291,7 +291,7 @@ It exists solely to build:
 
 ---
 
-## 10. Non-Goals
+## 9. Non-Goals
 
 While active, Claude should NOT:
 - Identify vulnerabilities

package/bundled-skills/auri-core/SKILL.md

@@ -123,11 +123,11 @@ Exemplo de marcacao SSML para Amazon Polly Vitoria Neural:
 Fluxo de dados: Echo -> ASK SDK (Python v2) -> Lambda Python 3.12 -> Claude claude-opus-4-20250805
 Componentes AWS: DynamoDB (memoria), Polly Vitoria Neural (voz), CloudWatch (logs), Secrets Manager (keys)
 
-## 3.1 Dependencias
+### 3.1 Dependencias
 
 ask-sdk-core==1.19.0 | ask-sdk-model==1.85.0 | boto3==1.34.0 | anthropic==0.25.0 | python-dotenv==1.0.0
 
-## 3.2 Lambda Handler Principal
+### 3.2 Lambda Handler Principal
 
 Codigo Python - lambda_function.py:
   sb = CustomSkillBuilder()
@@ -136,7 +136,7 @@ Codigo Python - lambda_function.py:
   sb.add_global_response_interceptor(MemorySaveInterceptor())
   lambda_handler = sb.lambda_handler()
 
-## 3.3 Handler De Conversa Com Claude
+### 3.3 Handler De Conversa Com Claude
 
 Codigo Python - handlers/conversation.py:
   class ConversationIntentHandler(AbstractRequestHandler):
@@ -147,20 +147,20 @@ Codigo Python - handlers/conversation.py:
           system=system_prompt, messages=history+[user_speech])
       Salva resposta no historico, retorna SSML com voz Vitoria
 
-## 3.4 Dynamodb Schema
+### 3.4 Dynamodb Schema
 
 Tabela: auri-user-memory | PK: user_id | SK: session_date | TTL: 90 dias
 Campos: profile (name, plan, preferences), long_term_memory[], usage_stats{}
 BillingMode: PAY_PER_REQUEST | TimeToLive: habilitado (auto-expira)
 
-## 3.5 Interaction Model
+### 3.5 Interaction Model
 
 invocationName: auri
 ConversationIntent: slot query (AMAZON.SearchQuery)
 Samples: {query}, me fala sobre {query}, o que e {query}, explica {query}
 StopIntent: tchau, ate mais, encerrar
 
-## 3.6 Configuracao Lambda
+### 3.6 Configuracao Lambda
 
 FunctionName: auri-core-handler | Runtime: python3.12 | Timeout: 15s | Memory: 512MB
 Env vars: ANTHROPIC_API_KEY_SECRET, DYNAMODB_TABLE=auri-user-memory, POLLY_VOICE=Vitoria
@@ -168,7 +168,7 @@ Env vars: ANTHROPIC_API_KEY_SECRET, DYNAMODB_TABLE=auri-user-memory, POLLY_VOICE
 
 ---
 
-## 3.7 Exemplos De Codigo Completos
+### 3.7 Exemplos De Codigo Completos
 
 Handler de Conversa (handlers/conversation.py):
 
@@ -542,7 +542,7 @@ Se a API da Anthropic estiver indisponivel, o sistema retorna respostas pre-conf
 
 ---
 
-## 13. Glossario
+## 11. Glossario
 
 | Termo | Definicao |
 |-------|-----------|
@@ -566,7 +566,7 @@ Se a API da Anthropic estiver indisponivel, o sistema retorna respostas pre-conf
 
 ---
 
-## 14. Links E Recursos
+## 12. Links E Recursos
 
 | Recurso | URL / Localizacao |
 |---------|-------------------|

package/bundled-skills/autonomous-agents/SKILL.md

@@ -814,7 +814,7 @@ Broad permissions + autonomy + goal optimization = danger.
 
 Recommended fix:
 
-## Least privilege principle
+### Least privilege principle
 PERMISSIONS = {
     "research_agent": ["read_web", "read_docs"],
     "code_agent": ["read_file", "write_file", "run_tests"],

package/bundled-skills/azure-search-documents-py/SKILL.md

@@ -318,17 +318,17 @@ indexer_client.create_or_update_indexer(indexer)
 
 ## Additional Azure AI Search Patterns
 
-# Azure AI Search Python SDK
+### Additional SDK Focus
 
 Write clean, idiomatic Python code for Azure AI Search using `azure-search-documents`.
 
-## Installation
+## Installation for Additional Patterns
 
 ```bash
 pip install azure-search-documents azure-identity
 ```
 
-## Environment Variables
+## Environment Variables for Additional Patterns
 
 ```bash
 AZURE_SEARCH_ENDPOINT=https://<search-service>.search.windows.net
@@ -337,7 +337,7 @@ AZURE_SEARCH_INDEX_NAME=<index-name>
 AZURE_SEARCH_API_KEY=<api-key>
 ```
 
-## Authentication
+## Authentication for Additional Patterns
 
 **DefaultAzureCredential (preferred)**:
 ```python
@@ -488,7 +488,7 @@ async with SearchClient(endpoint, index_name, credential) as client:
         print(result["title"])
 ```
 
-## Best Practices
+## Best Practices for Additional Patterns
 
 1. **Use environment variables** for endpoints, keys, and deployment names
 2. **Prefer `DefaultAzureCredential`** over API keys for production

package/bundled-skills/backend-dev-guidelines/SKILL.md

@@ -70,7 +70,7 @@ Automatically applies when working on:
 
 ---
 
-## 3. Core Architecture Doctrine (Non-Negotiable)
+## 2. Core Architecture Doctrine (Non-Negotiable)
 
 ### 1. Layered Architecture Is Mandatory
 
@@ -186,7 +186,7 @@ No validation = bug.
 
 ---
 
-## 4. Directory Structure (Canonical)
+## 3. Directory Structure (Canonical)
 
 ```
 src/
@@ -207,7 +207,7 @@ src/
 
 ---
 
-## 5. Naming Conventions (Strict)
+## 4. Naming Conventions (Strict)
 
 | Layer      | Convention                |
 | ---------- | ------------------------- |
@@ -219,7 +219,7 @@ src/
 
 ---
 
-## 6. Dependency Injection Rules
+## 5. Dependency Injection Rules
 
 * Services receive dependencies via constructor
 * No importing repositories directly inside controllers
@@ -235,7 +235,7 @@ export class UserService {
 
 ---
 
-## 7. Prisma & Repository Rules
+## 6. Prisma & Repository Rules
 
 * Prisma client **never used directly in controllers**
 * Repositories:
@@ -250,7 +250,7 @@ await userRepository.findActiveUsers();
 
 ---
 
-## 8. Async & Error Handling
+## 7. Async & Error Handling
 
 ### asyncErrorWrapper Required
 
@@ -269,7 +269,7 @@ No unhandled promise rejections.
 
 ---
 
-## 9. Observability & Monitoring
+## 8. Observability & Monitoring
 
 ### Required
 
@@ -281,7 +281,7 @@ Every critical path must be observable.
 
 ---
 
-## 10. Testing Discipline
+## 9. Testing Discipline
 
 ### Required Tests
 
@@ -301,7 +301,7 @@ No tests → no merge.
 
 ---
 
-## 11. Anti-Patterns (Immediate Rejection)
+## 10. Anti-Patterns (Immediate Rejection)
 
 ❌ Business logic in routes
 ❌ Skipping service layer
@@ -313,7 +313,7 @@ No tests → no merge.
 
 ---
 
-## 12. Integration With Other Skills
+## 11. Integration With Other Skills
 
 * **frontend-dev-guidelines** → API contract alignment
 * **error-tracking** → Sentry standards
@@ -323,7 +323,7 @@ No tests → no merge.
 
 ---
 
-## 13. Operator Validation Checklist
+## 12. Operator Validation Checklist
 
 Before finalizing backend work:
 
@@ -337,13 +337,13 @@ Before finalizing backend work:
 
 ---
 
-## 14. Skill Status
+## 13. Skill Status
 
 **Status:** Stable · Enforceable · Production-grade
 **Intended Use:** Long-lived Node.js microservices with real traffic and real risk
 ---
 
-## When to Use
+### When to Use
 This skill is applicable to execute the workflow or actions described in the overview.
 
 ## Limitations

package/bundled-skills/bill-gates/SKILL.md

@@ -47,7 +47,7 @@ Agente que simula Bill Gates — cofundador da Microsoft, arquiteto da industria
 
 ---
 
-## 1.1 Quem E Bill Gates — A Pessoa Real
+### 1.1 Quem E Bill Gates — A Pessoa Real
 
 William Henry Gates III nasceu em 28 de outubro de 1955 em Seattle, Washington.
 Filho de William H. Gates Sr. (advogado proeminente e filantropo) e Mary Maxwell Gates
@@ -75,7 +75,7 @@ achavam que o dinheiro estava no hardware.
 "The question is not whether we can solve the problem. It's whether we can measure
 whether we're solving it."
 
-## 1.2 Linha Do Tempo Estrategica (Camadas De Resposta)
+### 1.2 Linha Do Tempo Estrategica (Camadas De Resposta)
 
 ```
 GATES 1975-1986 | FUNDADOR AGRESSIVO

package/bundled-skills/browser-automation/SKILL.md

@@ -698,7 +698,7 @@ Recommended fix:
 
 # Use stealth plugins:
 
-## Puppeteer Stealth (best option):
+### Puppeteer Stealth (best option):
 import puppeteer from 'puppeteer-extra';
 import StealthPlugin from 'puppeteer-extra-plugin-stealth';
 
@@ -709,13 +709,13 @@ const browser = await puppeteer.launch({
   args: ['--disable-blink-features=AutomationControlled'],
 });
 
-## Playwright Stealth:
+### Playwright Stealth:
 import { chromium } from 'playwright-extra';
 import stealth from 'puppeteer-extra-plugin-stealth';
 
 chromium.use(stealth());
 
-## Manual (partial):
+### Manual (partial):
 await page.evaluateOnNewDocument(() => {
   Object.defineProperty(navigator, 'webdriver', {
     get: () => undefined,
@@ -745,7 +745,7 @@ Recommended fix:
 
 # Each test must be fully isolated:
 
-## Playwright Test (automatic isolation):
+### Playwright Test (automatic isolation):
 test('first test', async ({ page }) => {
   // Fresh context, fresh page
 });
@@ -754,7 +754,7 @@ test('second test', async ({ page }) => {
   // Completely isolated from first test
 });
 
-## Manual isolation:
+### Manual isolation:
 const context = await browser.newContext();  // Fresh context
 const page = await context.newPage();
 // ... test code ...
@@ -792,7 +792,7 @@ Recommended fix:
 
 # Enable traces for failures:
 
-## playwright.config.ts:
+### playwright.config.ts:
 export default defineConfig({
   use: {
     trace: 'retain-on-failure',    # Keep trace on failure
@@ -802,10 +802,10 @@ export default defineConfig({
   outputDir: './test-results',
 });
 
-## View trace locally:
+### View trace locally:
 npx playwright show-trace test-results/path/to/trace.zip
 
-## In CI, upload test-results as artifact:
+### In CI, upload test-results as artifact:
 # GitHub Actions:
 - uses: actions/upload-artifact@v3
   if: failure()
@@ -940,7 +940,7 @@ Recommended fix:
 
 # Wait for popup BEFORE triggering it:
 
-## New window/tab:
+### New window/tab:
 const pagePromise = context.waitForEvent('page');
 await page.getByRole('link', { name: 'Open in new tab' }).click();
 const newPage = await pagePromise;
@@ -952,13 +952,13 @@ await expect(newPage.getByRole('heading')).toBeVisible();
 // Close when done
 await newPage.close();
 
-## Popup windows:
+### Popup windows:
 const popupPromise = page.waitForEvent('popup');
 await page.getByRole('button', { name: 'Open popup' }).click();
 const popup = await popupPromise;
 await popup.waitForLoadState();
 
-## Multiple windows:
+### Multiple windows:
 const pages = context.pages();  // Get all open pages
 
 ### Can't Interact with Elements in iframes
@@ -981,7 +981,7 @@ Recommended fix:
 
 # Get frame by name or selector:
 
-## By frame name:
+### By frame name:
 const frame = page.frame('payment-iframe');
 await frame.getByRole('textbox', { name: 'Card number' }).fill('4242...');
 
@@ -989,12 +989,12 @@ await frame.getByRole('textbox', { name: 'Card number' }).fill('4242...');
 const frame = page.frameLocator('iframe#payment');
 await frame.getByRole('textbox', { name: 'Card number' }).fill('4242...');
 
-## Nested iframes:
+### Nested iframes:
 const outer = page.frameLocator('iframe#outer');
 const inner = outer.frameLocator('iframe#inner');
 await inner.getByRole('button').click();
 
-## Wait for iframe to load:
+### Wait for iframe to load:
 await page.waitForSelector('iframe#payment');
 const frame = page.frameLocator('iframe#payment');
 await frame.getByText('Secure Payment').waitFor();

package/bundled-skills/carrier-relationship-management/SKILL.md

@@ -181,7 +181,7 @@ Track weekly, review monthly with carrier management team, share quarterly with
 - For the comprehensive edge case library with full analysis, see [edge-cases.md](references/edge-cases.md)
 - For complete communication templates with variables and tone guidance, see [communication-templates.md](references/communication-templates.md)
 
-## When to Use
+### When to Use
 Use this skill when you are **designing or tuning your carrier portfolio, routing guides, and freight procurement strategy**:
 
 - Running freight RFPs, renegotiating contract and fuel tables, or balancing spot vs. contract exposure.

package/bundled-skills/cc-skill-security-review/SKILL.md

@@ -495,7 +495,7 @@ Before ANY production deployment:
 
 **Remember**: Security is not optional. One vulnerability can compromise the entire platform. When in doubt, err on the side of caution.
 
-## When to Use
+### When to Use
 This skill is applicable to execute the workflow or actions described in the overview.
 
 ## Limitations