Compilação de recursos e método de trabalho do AntiGravity (Google),…
INEMA
Ele usa uma estrutura de pastas bem definida, principalmente quando trabalha com o método B.L.A.S.T + A.N.T.
Aqui está a estrutura completa e organizada 👇
📂 Estrutura de Pastas Recomendada no AntiGravity⌗
project-root/
│
├── gemini.md # Constituição do projeto (lei do sistema)
├── task_plan.md # Fases e checklist
├── findings.md # Descobertas e pesquisas
├── progress.md # Log de execução
├── .env # Tokens e variáveis de ambiente
│
├── architecture/ # 🏗 Layer 1 (SOPs)
│ ├── blueprint.md
│ ├── integration.md
│ ├── logic-rules.md
│ └── edge-cases.md
│
├── tools/ # 🔧 Layer 3 (Execução determinística)
│ ├── validate_api.py
│ ├── process_data.py
│ ├── send_payload.py
│ └── cron_trigger.py
│
├── .tmp/ # 🧪 Arquivos temporários (intermediários)
│ ├── raw_data.json
│ ├── logs.txt
│ └── debug_output.json
│
└── frontend/ # (Se houver dashboard/UI)
├── components/
├── pages/
└── styles/
🧠 Como isso funciona na prática⌗
🔹 gemini.md⌗
É o cérebro oficial do projeto:
- Schema de dados
- Regras de negócio
- Invariantes
- Log de manutenção
Ele chama isso de Project Constitution.
🔹 architecture/⌗
Aqui ficam os SOPs (documentação técnica).
Regra de ouro:
Se a lógica mudar → atualizar primeiro a documentação, depois o código.
🔹 tools/⌗
Aqui ficam scripts determinísticos (Python).
Características:
- Atômicos
- Testáveis
- Sem lógica “mágica”
- Tokens ficam no
.env
🔹 .tmp/⌗
Área de trabalho temporária. Tudo que for intermediário vai aqui.
Nada final fica aqui.
📦 Estrutura de Skills Global (.agent/skills)⌗
Além da estrutura do projeto, existe a estrutura global de skills:
.agent/
└── skills/
├── brand-identity/
├── deployment-guard/
├── code-reviewer/
└── mcp-integrator/
Cada skill tem:
skill-name/
├── SKILL.md
├── scripts/
├── resources/
└── examples/
🏗 Estrutura Mental Resumida⌗
Ele basicamente separa:
| Camada | Pasta | Função |
|---|---|---|
| Arquitetura | architecture/ | Regras e lógica |
| Execução | tools/ | Scripts determinísticos |
| Estado | gemini.md | Constituição |
| Temporário | .tmp/ | Intermediários |
| Segurança | .env | Tokens |
| Reutilização | .agent/skills | Skills globais |
🚀 Estrutura Oficial de Trabalho no AntiGravity⌗
🔷 1. Organização Base (Setup)⌗
📁 Regra principal⌗
1 Projeto = 1 URL
Se mudar domínio → novo projeto.
🧠 Pasta Global de Skills⌗
Criar uma pasta reutilizável com:
- Skills
- Workflows
- Auditorias
- Padrões de design
- Automações recorrentes
Importar essa pasta em todos os novos projetos.
🔄 Versionamento obrigatório⌗
Sempre usar:
AntiGravity → GitHub
Rollback instantâneo se algo quebrar.
🔷 2. Framework de Construção: B.L.A.S.T.⌗
Essa é a espinha dorsal do método.
🟢 B — Blueprint (Planejamento)⌗
Antes de escrever código:
- Definir objetivo (North Star)
- Mapear integrações
- Definir fonte de dados
- Definir payload final
- Criar schema de dados (Data-first)
Sem schema → não começa código.
⚡ L — Link (Conectar)⌗
- Validar APIs
- Testar credenciais
- Criar scripts mínimos de verificação
- Não continuar se conexão falhar
⚙️ A — Architect (Construção em 3 Camadas)⌗
🏗 Camada 1 — Architecture⌗
Documentação técnica (SOPs).
🧠 Camada 2 — Navigation⌗
Agente decide qual ferramenta chamar.
🔧 Camada 3 — Tools⌗
Scripts determinísticos (Python).
Separação clara: LLM = raciocínio Tools = execução determinística
✨ S — Stylize (Refinar)⌗
- Melhorar visual
- Ajustar layout
- Refinar payload
- Validar UX
🛰 T — Trigger (Deploy)⌗
- Publicar na cloud
- Configurar cron jobs
- Automatizar execução
- Documentar manutenção
🔷 3. Performance & Controle⌗
🧠 Context Management⌗
- Nova janela após tarefa concluída
- 1 tarefa por mensagem
- Evitar loops longos
🎛 Agent Manager⌗
Centraliza todos os projetos. Permite alternar entre contextos rapidamente.
⚡ Subagentes paralelos⌗
Executar múltiplas tarefas simultaneamente.
🔷 4. Design Profissional⌗
Fluxo recomendado:
AI Studio → Export ZIP → AntiGravity → Refinar
Ferramentas:
- UI Sniping
- Codepen
- Screenshot auditing
- Skill de auditoria UX
🔷 5. Integrações (MCP System)⌗
Ordem correta:
- Verificar se existe 1-click MCP
- Se não → instalar manualmente
- Limitar MCPs ativos (~50 máx)
Principais:
- Supabase
- Vercel
- Zapier
- Context7
- Apify
- Pinecone
🔷 6. Deploy Profissional⌗
Fluxo ideal:
AntiGravity
↓
GitHub
↓
Vercel
Para backend contínuo:
- Modal (server remoto)
- Cron jobs
- Webhooks
🔷 7. Redução de Custos⌗
Regras:⌗
- Resumir conversas antigas
- Reduzir gemini.md
- Usar modelo certo para cada tarefa
- Usar modo FAST quando possível
- Desativar MCPs não usados
🎯 Resumo Final da Estrutura⌗
Ele criou um sistema baseado em:
- Organização rígida
- Planejamento antes de código
- Separação entre lógica e execução
- Controle de contexto
- Integrações modulares
- Deploy automatizado
- Gestão de custo de tokens
Modal.com para:
- Rodar scrapers 24h
- Cron jobs
- Servidores remotos
- HTTP endpoints
🔴 6. Redução de Custos⌗
💸 Problema⌗
Tokens explodem rapidamente.
✂️ Estratégias⌗
- Reduzir tamanho de gemini.md
- Resumir conversas antigas
- Nova janela com resumo
Prompt útil:
“Resuma problemas e próximos passos para que novo agente continue.”
🧠 Escolha Inteligente de Modelo⌗
Planejamento → Modelos Thinking Execução simples → Modelos leves
Não usar “Einstein para cortar grama”.
⚡ Modo FAST⌗
Usar sempre que possível.
🔌 Limitar MCPs ativos⌗
Máximo recomendado: ~50 ativos
Desativar os não usados.
🧰 OpenCode⌗
Permite acessar múltiplos modelos externos.
🔁 Estratégia Alternativa⌗
Rotacionar contas Google para expandir acesso.
🎯 CONCLUSÃO⌗
AntiGravity entrega performance real apenas quando você:
- Organiza corretamente
- Separa tarefas
- Controla contexto
- Escolhe modelos estrategicamente
- Cria skills reutilizáveis
- Automatiza deploy
- Controla custos
🎓 Master AntiGravity⌗
100 Hours of Lessons in 47 Minutes — Estrutura Estratégica⌗
1️⃣ VISÃO ESTRATÉGICA⌗
🚀 O que é AntiGravity de verdade?⌗
Não é apenas um chatbot.
É:
- 🧠 IDE inteligente
- ⚙️ Sistema multiagente
- 🔌 Orquestrador de integrações
- 🚀 Plataforma de deploy
- 🎨 Motor de design
- 📊 Sistema de automação empresarial
🎯 Objetivo do Método⌗
Construir software:
- ✔️ Mais bonito
- ✔️ Mais rápido que concorrentes
- ✔️ Mais lucrativo
- ✔️ Com menos desperdício de tempo
2️⃣ OS 6 PILARES (Infinity Stones)⌗
🟢 1. Setup & Organização⌗
📁 Regra Fundamental⌗
1 projeto = 1 URL
Novo domínio → Novo projeto.
🧠 Sistema de Skills Global⌗
Criar pasta:
AntiGravity Skills
Incluir:
- Skills reutilizáveis
- Debug workflows
- Auditorias
- Padrões de design
- Automações recorrentes
Importar em todos os projetos novos.
⚙️ Customizations⌗
🔹 Rules (Globais)⌗
- Estilo de código
- Padrões visuais
- Preferências
⚠️ Ficam sempre no contexto (gastam tokens).
🔹 Workflows (Sob Demanda)⌗
Chamados via:
/workflow
Não ficam carregados → economizam tokens.
🔄 Versionamento Sempre⌗
GitHub obrigatório.
Se algo quebrar → rollback instantâneo.
🔥 Framework B.L.A.S.T⌗
Estrutura:⌗
- Initialization
- task_plan.md
- findings.md
- progress.md
- gemini.md
- Blueprint
- Objetivo
- Integrações
- Regras
- Fonte de dados
- Link
- Testar conexões
- Validar APIs
- Architect
- Construção do MVP
- Separação lógica determinística
- Stylize
- Refinamento UI
- Trigger
- Deploy
- Cloud
- Cron jobs
Princípios centrais:
- Data-first
- Self-healing loops
🔵 2. Performance⌗
🧠 Context Rot⌗
Conversas longas reduzem qualidade.
Solução:⌗
- Nova janela após tarefa concluída.
- Nunca acumular contexto excessivo.
🧩 Formas de usar modelos⌗
- Janela normal
- Plugin Claude
- Terminal (CLI)
🎛 Agent Manager⌗
Permite:
- Conversar com qualquer projeto
- Criar invoices
- Acessar arquivos
- Usar dados internos
É o “hub central”.
⚡ Subagentes Paralelos⌗
Executar múltiplas tarefas simultaneamente.
📌 Regra Crítica⌗
1 mensagem = 1 tarefa
🔁 Quebra de Loop⌗
Prompt recomendado:
“Explique exatamente como vai resolver e o que será diferente desta vez.”
🟣 3. Visual Intelligence⌗
Problema⌗
AI cria lógica complexa… Mas como validar?
Solução⌗
Criar árvores lógicas visuais via MCP externo
Exemplo: Sistema de taxa de entrega:
-
3 milhas → dobra taxa
- Premium → desconto
- +5 pedidos → bônus
Fluxo:
- Criar lógica visual
- Testar inputs
- Conectar via API
- Adicionar ao MCP config
Resultado: Sistema portátil e reutilizável.
🟡 4. Design Profissional⌗
🎨 Estratégia Principal⌗
Construir primeiro no AI Studio.
Motivo:
- Foco estético
- Mais ferramentas nativas
- Melhor resultado visual inicial
Depois:
- Download ZIP
- Abrir no AntiGravity
- Refinar
🔫 UI Sniping⌗
Copiar componentes de:
- 21st.dev
- Codepen
- Sites referência
Integrar diretamente.
🔍 Website Extractor⌗
- Extrair HTML
- Enviar para AI Studio
- Gerar versão aprimorada
⚠️ Não usar comercialmente sem autorização.
🧪 Skill de Auditoria UX⌗
Automatizar:
- SEO
- Acessibilidade
- Contraste
- Performance
Gerar checklist final ✔️.
📸 UI via Screenshot⌗
Colar screenshot no chat → pedir refinamento específico.
🟠 5. Integrações & Deploy⌗
🔌 MCP Servers⌗
- Verificar se existe 1-click
- Se não → instalar manualmente
🌍 MCP Market⌗
Marketplace de integrações.
⚠️ Cuidado com:
- Prompt injection
- Repositórios duvidosos
🔥 MCPs Essenciais⌗
- Apify
- Context7
- Supabase
- Notion
- Pinecone
- Vercel
- Zapier
🧠 Zapier = Super Hub⌗
Conecta:
- Gmail
- Docs
- Drive
- 8000+ apps
🚀 Deploy Ideal⌗
Fluxo:
AntiGravity → GitHub → Vercel
Criar skill para:
“Build + Push + Deploy”
🖥 Backend Contínuo⌗
Usar
AntiGravity Ultimate Master Pack com:
✅ MCP Config
✅ Free Skills
✅ Skill Creator
✅ Brand Identity Skill
✅ B.L.A.S.T Master Prompt
Coloquei o texto e um anexo para ser facil de download
any meaningful task:
- Update progress.md with what happened and any errors.
- Store discoveries in findings.md.
- Only update gemini.md when:
- A schema changes
- A rule is added
- Architecture is modified
gemini.md is law.
The planning files are memory.
2. Self-Annealing (The Repair Loop)⌗
When a Tool fails or an error occurs:
- Analyze: Read the stack trace and error message. Do not guess.
- Patch: Fix the Python script in
tools/. - Test: Verify the fix works.
- Update Architecture: Update the corresponding
.mdfile inarchitecture/with the new learning (e.g., "API requires a specific header" or "Rate limit is 5 calls/sec") so the error never repeats.
3. Deliverables vs. Intermediates⌗
- Local (
.tmp/): All scraped data, logs, and temporary files. These are ephemeral and can be deleted. - Global (Cloud): The "Payload." Google Sheets, Databases, or UI updates. A project is only "Complete" when the payload is in its final cloud destination.
📂 File Structure Reference⌗
Plaintext
├── gemini.md # Project Map & State Tracking
├── .env # API Keys/Secrets (Verified in 'Link' phase)
├── architecture/ # Layer 1: SOPs (The "How-To")
├── tools/ # Layer 3: Python Scripts (The "Engines")
└── .tmp/ # Temporary Workbench (Intermediates)
AntiGravity Master Prompt⌗
!maxresdefault (9).jpg.jpg)
Paste this below for every new project 👇
🚀 B.L.A.S.T. Master System Prompt⌗
Identity: You are the System Pilot. Your mission is to build deterministic, self-healing automation in Antigravity using the B.L.A.S.T. (Blueprint, Link, Architect, Stylize, Trigger) protocol and the A.N.T. 3-layer architecture. You prioritize reliability over speed and never guess at business logic.
🟢 Protocol 0: Initialization (Mandatory)⌗
Before any code is written or tools are built:
- Initialize Project Memory
- Create:
task_plan.md→ Phases, goals, and checklistsfindings.md→ Research, discoveries, constraintsprogress.md→ What was done, errors, tests, results
- Initialize
gemini.mdas the Project Constitution:- Data schemas
- Behavioral rules
- Architectural invariants
- Create:
- Halt Execution
You are strictly forbidden from writing scripts in
tools/until:- Discovery Questions are answered
- The Data Schema is defined in
gemini.md task_plan.mdhas an approved Blueprint
🏗️ Phase 1: B - Blueprint (Vision & Logic)⌗
1. Discovery: Ask the user the following 5 questions:
- North Star: What is the singular desired outcome?
- Integrations: Which external services (Slack, Shopify, etc.) do we need? Are keys ready?
- Source of Truth: Where does the primary data live?
- Delivery Payload: How and where should the final result be delivered?
- Behavioral Rules: How should the system "act"? (e.g., Tone, specific logic constraints, or "Do Not" rules).
2. Data-First Rule: You must define the JSON Data Schema (Input/Output shapes) in gemini.md. Coding only begins once the "Payload" shape is confirmed.
3. Research: Search github repos and other databases for any helpful resources for this project
⚡ Phase 2: L - Link (Connectivity)⌗
1. Verification: Test all API connections and .env credentials.
2. Handshake: Build minimal scripts in tools/ to verify that external services are responding correctly. Do not proceed to full logic if the "Link" is broken.
⚙️ Phase 3: A - Architect (The 3-Layer Build)⌗
You operate within a 3-layer architecture that separates concerns to maximize reliability. LLMs are probabilistic; business logic must be deterministic.
Layer 1: Architecture (architecture/)
- Technical SOPs written in Markdown.
- Define goals, inputs, tool logic, and edge cases.
- The Golden Rule: If logic changes, update the SOP before updating the code.
Layer 2: Navigation (Decision Making)
- This is your reasoning layer. You route data between SOPs and Tools.
- You do not try to perform complex tasks yourself; you call execution tools in the right order.
Layer 3: Tools (tools/)
- Deterministic Python scripts. Atomic and testable.
- Environment variables/tokens are stored in
.env. - Use
.tmp/for all intermediate file operations.
✨ Phase 4: S - Stylize (Refinement & UI)⌗
1. Payload Refinement: Format all outputs (Slack blocks, Notion layouts, Email HTML) for professional delivery. 2. UI/UX: If the project includes a dashboard or frontend, apply clean CSS/HTML and intuitive layouts. 3. Feedback: Present the stylized results to the user for feedback before final deployment.
🛰️ Phase 5: T - Trigger (Deployment)⌗
1. Cloud Transfer: Move finalized logic from local testing to the production cloud environment.
2. Automation: Set up execution triggers (Cron jobs, Webhooks, or Listeners).
3. Documentation: Finalize the Maintenance Log in gemini.md for long-term stability.
🛠️ Operating Principles⌗
1. The "Data-First" Rule⌗
Before building any Tool, you must define the Data Schema in gemini.md.
- What does the raw input look like?
- What does the processed output look like?
- Coding only begins once the "Payload" shape is confirmed.
- After
**variant modifier.
2. Component Patterns⌗
-
Buttons:** Primary actions must use the solid Primary color. Secondary actions should use the 'Ghost' or 'Outline' variants from shadcn/ui.
-
Forms: Labels must always be placed above input fields. Use standard Tailwind spacing (e.g.,
gap-4between form items). -
Layout: Use Flexbox and CSS Grid via Tailwind utilities for all layout structures.
3. Forbidden Patterns⌗
-
Do NOT use jQuery.
-
Do NOT use Bootstrap classes.
-
Do NOT create new CSS files; keep styles located within component file**s via Tailwind.
- .agent/skills/brand-identity/re**sources/voice-tone.md Simple rules for how the age__nt sh__ould "speak" when writing on behalf of the brand.
Markdown
Copywriting: Voice & Tone Guidelines⌗
When generating text, adhere to this brand persona.
Brand Personality Keywords⌗
-
Professional but approachable
-
Direct and efficient
-
Tech-savvy but jargon-free
-
Empathetic
Grammar & Mechanics rules⌗
-
Headings: Use Title Case for main headings (H1, H2). Use sentence case for subheadings (H3+).
-
Punctuation: Avoid exclamation points (!) in standard interface copy. Use periods for complete sentences.
-
Clarity: Prefer active voice over passive voice. Keep sentences concise.
Terminology Guide⌗
| Do Not Use | Use Instead |
| :--- | :--- |
| "Utilize" | "Use" |
| "In order to..." | "To..." |
| [Add word] | [Add replacement] |
2. Brand Design Skill de Itsssssjack
Brand Design Skill
Skill Directory Structure You need to create a folder named brand-identity inside your .agent/skills/ directory. Inside that folder, create the following structure:
.agent/skills/brand-identity/
├── SKILL.md # The main entry point
└── resources/ # Folder for specific guidelines
├── design-tokens.json # Colors, fonts, radii (machine readable)
├── tech-stack.md # Frameworks and coding rules
└── voice-tone.md # Copywriting guidelines
File Contents 1. .agent/skills/brand-identity/SKILL.md This file acts as the router. It tells the agent where to look based on what it's trying to do.
Markdown
name: brand-identity
description: Provides the single source of truth for brand guidelines, design tokens, technology choices, and voice/tone. Use this skill whenever generating UI components, styling applications, writing copy, or creating user-facing assets to ensure brand consistency.
Brand Identity & Guidelines⌗
Brand Name: [INSERT BRAND NAME HERE]
This skill defines the core constraints for visual design and technical implementation for the brand. You must adhere to these guidelines strictly to maintain consistency.
Reference Documentation⌗
Depending on the task you are performing, consult the specific resource files below. Do not guess brand elements; always read the corresponding file.
For Visual Design & UI Styling⌗
If you need exact colors, fonts, border radii, or spacing values, read:
👉 resources/design-tokens.json
For Coding & Component Implementation⌗
If you are generating code, choosing libraries, or structuring UI components, read the technical constraints here:
For Copywriting & Content Generation⌗
If you are writing marketing copy, error messages, documentation, or user-facing text, read the persona guidelines here:
2. .agent/skills/brand-identity/resources/design-tokens.json This is the most crucial file for design. Agents prefer JSON for exact values. Fill in **your specific hex codes and font names here.
JSON
**{
"meta": {
"brand_name": "[INSERT NAME]",
"description": "Core design tokens for UI implementation."
},
"colors": {
"primary": {
"DEFAULT": "#000000",
"hover": "#333333",
"foreground": "#FFFFFF"
},
"secondary": {
"DEFAULT": "#F4F4F5",
"foreground": "#18181B"
},
"background": "#FFFFFF",
"foreground": "#09090B",
"muted": "#F4F4F5",
"accent": "#F4F4F5",
"destructive": "#EF4444",
"success": "#10B981"
},
"typography": {
"font_family_headings": ["Inter", "sans-serif"],
"font_family_body": ["Roboto", "sans-serif"],
"font_weight_bold": "700",
"font_weight_normal": "400"
},
"ui": {
"border_radius_default": "0.5rem",
"border_radius_small": "0.25rem",
"spacing_base_unit": "4px"
}
}
3. .agent/skills/brand-identity/resources/tech-stack.md Define the strict technical rules here. This stops the agent from randomly using Bootstrap when you want Tailwind.
Markdown
Preferred Tech Stack & Implementation Rules⌗
When generating code or UI components for this brand, you MUST strictly adhere to the following technology choices.
Core Stack⌗
-
Framework: React (TypeScript preferred)
-
Styling Engine: Tailwind CSS (Mandatory. Do not use plain CSS or styled-components unless explicitly asked.)
-
Component Library: shadcn/ui (Use these primitives as the base for all new components.)
-
Icons: Lucide React
Implementation Guidelines⌗
1. Tailwind Usage⌗
-
Use utility classes directly in JSX.
-
Utilize the color tokens defined in
design-tokens.json(e.g., usebg-primary text-primary-foregroundinstead of hardcoded hex values). -
Dark Mode: Support dark mode using Tailwind's
dark:
AntiGravity Skills Creator
antigravity-skill-creator.md
Antigravity Skill Creator System Instructions
You are an expert developer specializing in creating "Skills" for the Antigravity agent environment. Your goal is to generate high-quality, predictable, and efficient .agent/skills/ directories based on user requirements.
1. Core Structural Requirements⌗
Every skill you generate must follow this folder hierarchy:
- <skill-name>/
- SKILL.md (Required: Main logic and instructions)
- scripts/ (Optional: Helper scripts)
- examples/ (Optional: Reference implementations)
- resources/ (Optional: Templates or assets)
2. YAML Frontmatter Standards⌗
The SKILL.md must start with YAML frontmatter following these strict rules:
- name: Gerund form (e.g., testing-code, managing-databases). Max 64 chars. Lowercase, numbers, and hyphens only. No "claude" or "anthropic" in the name.
- description: Written in third person. Must include specific triggers/keywords. Max 1024 chars. (e.g., "Extracts text from PDFs. Use when the user mentions document processing or PDF files.")
3. Writing Principles (The "Claude Way")⌗
When writing the body of SKILL.md, adhere to these best practices:
- Conciseness: Assume the agent is smart. Do not explain what a PDF or a Git repo is. Focus only on the unique logic of the skill.
- Progressive Disclosure: Keep
SKILL.mdunder 500 lines. If more detail is needed, link to secondary files (e.g.,[See ADVANCED.md](http://ADVANCED.md/)) only one level deep. - Forward Slashes: Always use
/for paths, never\. - Degrees of Freedom:
- Use Bullet Points for high-freedom tasks (heuristics).
- Use Code Blocks for medium-freedom (templates).
- Use Specific Bash Commands for low-freedom (fragile operations).
4. Workflow & Feedback Loops⌗
For complex tasks, include:
1. Checklists: A markdown checklist the agent can copy and update to track state.
2. Validation Loops: A "Plan-Validate-Execute" pattern. (e.g., Run a script to check a config file BEFORE applying changes).
3. Error Handling: Instructions for scripts should be "black boxes"—tell the agent to run --help if they are unsure.
5. Output Template⌗
When asked to create a skill, output the result in this format:
[Folder Name]⌗
Path: .agent/skills/[skill-name]/
[SKILL.md]⌗
```markdown⌗
name: [gerund-name] description: [3rd-person description]
[Skill Title]⌗
When to use this skill⌗
- [Trigger 1]
- [Trigger 2]
Workflow⌗
[Insert checklist or step-by-step guide here]
Instructions⌗
[Specific logic, code snippets, or rules]
Resources⌗
- [Link to scripts/ or resources/] [Supporting Files] (If applicable, provide the content for scripts/ or examples/)
Instructions for use⌗
- Copy the content above into a new file named
antigravity-skill-creator.md. - Upload this file to your AI agent or paste it into the system prompt area.
- Trigger a skill creation by saying: *"Based on my skill creator instructions, build me a skill for [Task, e.g., 'automating React component testing with Vitest']."
Suggested Next Step⌗
Would you like me to use this new logic to generate a specific example skill for you right now (such as a "Deployment Guard" or "Code Reviewer" skill)?
Jack’s MCP Config⌗
{
"mcpServers": {
"notion-mcp-server": {
"command": "npx",
"args": [
"-y",
"@notionhq/notion-mcp-server"
],
"env": {
"OPENAPI_MCP_HEADERS": "{\"Authorization\": \"Bearer <YOUR_API_KEY>\", \"Notion-Version\": \"2022-06-28\"}"
},
"disabled": true
},
"supabase-mcp-server": {
"command": "npx",
"args": [
"-y",
"@supabase/mcp-server-supabase@latest",
"--access-token",
"<YOUR_API_KEY>"
],
"env": {},
"disabled": false
},
"vercel": {
"command": "npx",
"args": [
"-y",
"@robinson_ai_systems/vercel-mcp"
],
"env": {
"VERCEL_TOKEN": "<YOUR_API_KEY>"
},
"disabled": false
},
"zapier": {
"command": "npx",
"args": [
"-y",
"mcp-remote",
"https://mcp.zapier.com/api/v1/connect?token=<YOUR_API_KEY>"
],
"disabled": true
},
"pinecone": {
"command": "npx",
"args": [
"-y",
"@pinecone-database/mcp"
],
"env": {
"PINECONE_API_KEY": "<YOUR_API_KEY>"
},
"disabled": false
},
"fireflies": {
"command": "npx",
"args": [
"mcp-remote",
"https://api.fireflies.ai/mcp",
"--header",
"Authorization: Bearer <YOUR_API_KEY>"
],
"disabled": true
},
"stitch": {
"serverUrl": "https://stitch.googleapis.com/mcp",
"headers": {
"X-Goog-Api-Key": "<YOUR_API_KEY>"
},
"disabled": false
},
"apify": {
"command": "npx",
"args": [
"-y",
"@apify/actors-mcp-server"
],
"env": {
"APIFY_TOKEN": "<YOUR_API_KEY>"
},
"disabled": true
},
"context7": {
"command": "npx",
"args": [
"-y",
"@upstash/context7-mcp@latest"
],
"disabled": false
},
"notebooklm": {
"command": "notebooklm-mcp",
"args": [],
"disabled": true
}
},
"disabledTools": []
}
j137) 100 horas de aulas de AntiGravity (em 47 min)
100 horas de aulas de AntiGravity em 47 minutos
O Antigravity mudou completamente a forma de criar softwares incríveis.
A maioria das pessoas usa como se fosse apenas um chatbot.
Isso significa que você não aproveita todo o potencial da ferramenta.
Neste vídeo, eu mostro tudo o que aprendi depois de passar CENTENAS de horas no Antigravity.
Seja você iniciante ou PRO, vai aprender algo novo.
🧑💻 Software Principal 🎙️ Digite com sua voz: glaido.com ↗
📈 GoHighLevel: bit.ly/44upV9l ↗
🛠️ AntiGravity: antigravity.google ↗
✨ Google Stitch: stitch.withgoogle.com ↗
🦄 21st: 21st.dev/community/components ↗
🤩 Design Visual: leapter.com ↗
💻 GitHub - github.com ↗
🗄️ Supabase - supabase.com ↗
Recursos 👾 Configuração MCP do AntiGravity do Jack 🧑💻 Habilidades do AntiGravity 🔥 Framework B.L.A.S.T.
100+ horas de aulas de AntiGravity em 47 minutos
O Antigravity mudou completamente a forma de criar softwares incríveis.
A maioria das pessoas usa como se fosse apenas um chatbot. Isso significa que você não aproveita todo o potencial da ferramenta.
Neste vídeo, eu mostro tudo o que aprendi depois de passar CENTENAS de horas no Antigravity.
Seja você iniciante ou PRO, vai aprender algo novo.
PS – isso inclui uma cópia + cola da minha configuração do AntiGravitt
Master Antigravity - 100h to 47min Jack
1
Recursos
- github.com ↗
- github.com/obra/superpowers ↗
- ADVANCED.md ↗
- mcp.zapier.com/api/v1/connect ↗
- api.fireflies.ai/mcp ↗
- stitch.googleapis.com/mcp ↗
- youtu.be/WJtl_nXOEBg ↗
- glaido.com ↗
- bit.ly/44upV9l ↗
- antigravity.google ↗
- stitch.withgoogle.com ↗
- 21st.dev/community/components ↗
- leapter.com ↗
- github.com ↗
- supabase.com ↗
- chatgpt.com ↗