Escolhendo um Modelo Claude
O trade-off entre Opus/Sonnet/Haiku e como escolher — sem adivinhar.
Tokens, Contexto e Preços
Conte tokens corretamente, max_tokens vs. janela de contexto e estimativa de custo.
Sua Primeira Chamada à API
Instale o SDK, autentique-se e faça uma chamada real ao Claude em Python, TypeScript ou cURL.
Streaming e Conversas Multi-Turno
Faça streaming de respostas e gerencie o histórico em uma API stateless.
Uso de Ferramentas / Function Calling
Defina ferramentas, execute o loop agêntico e trate os resultados.
Visão, PDF e Entrada de Arquivos
Envie imagens e documentos, e reutilize uploads por file_id.
Saída Estruturada
Obtenha JSON em conformidade com o schema, no qual você pode confiar — não na base do prompt-e-reza.
Raciocínio Estendido & Esforço
Raciocínio adaptativo e o controle de esforço — escolhendo a profundidade conforme a carga de trabalho.
Effort tuning: 5 levels, model defaults, and the cache trap
The July 2026 effort parameter has five levels — low, medium, high, xhigh, max — and lives inside output_config, not at the top level. It affects tool calls (not just thinking), invalidates prompt cache when changed mid-conversation, and each model has a different recommended starting point. The practical tuning guide.
The advisor tool: Sonnet does the work, Fable does the thinking
The advisor server-tool (beta advisor-tool-2026-03-01) lets a fast executor model consult a stronger model mid-turn. Full field reference, the token-accounting trap most integrators miss, the model pairing matrix, and the exact prompts Anthropic ships in production.
Cache de Prompt e Otimização de Custo
Reaproveite um prefixo de prompt estável para reduzir custo e latência — e a pegadinha que quebra tudo.
Mid-Conversation System Messages
Steer a long agent session without blowing up the prompt cache — the new system role you can append after any user turn.
Mid-Conversation Tool Changes
Add and retire tools between turns without invalidating the prompt cache — the Opus 5 beta that fixes the biggest hidden tax on long agent sessions.
Construindo Agentes na API
Chamada única vs workflow vs agente personalizado — o teste de decisão e o design do loop.
Agentes Gerenciados
Loop de agente hospedado pela Anthropic: agentes vs sessões, environments, memória, vaults e deployments agendados (cron).
Managed Agents Memory Stores
The July 2026 beta — server-side persistent memory that mounts into a Managed Agents sandbox, with versioning, audit trail, and redaction.
Cowork e Times de Agentes
O espaço de trabalho agêntico no desktop e a colaboração multiagente como produtos.
MCP e Conexão a Ferramentas
Declare servidores MCP nas requisições e onde o MCP se encaixa em relação a outros tipos de ferramentas.
MCP 2026-07-28: The Stateless Spec
The biggest MCP revision since launch: the session is gone, every request self-describes, and any server instance can answer any call. What actually changed on July 28, 2026, and how to migrate without breaking today's servers.
MCP Apps: Interactive UIs Inside a Tool Call
MCP Apps (SEP-1865) lets an MCP server ship an HTML UI that the client renders inside a sandboxed iframe — with every action still routed through the standard JSON-RPC audit path. What the spec actually says, what to build with it, and what to refuse to build.
MCP Tasks: Long-Running Work Without the Session
The 2026-07-28 stateless MCP spec killed the session — so how do you model a 20-minute CI job, a batch import, or a human approval gate? With the official Tasks extension (SEP-2663). Full lifecycle, wire protocol, migration from the old experimental Tasks API, and the polling/notification patterns that actually work in production.
Playwright MCP: The Deep Practical Guide (2026)
Microsoft's Playwright MCP is now the #1 MCP server in the world — ahead of GitHub and Figma. The bit almost nobody teaches: the default accessibility-snapshot mode is not just faster than vision, it's a completely different automation paradigm. Profiles, opt-in caps, browser-extension mode that attaches to your real Chrome, the token-cost fight against Playwright-as-a-Skill, and the security line you must not cross.
Segurança, Recusas e Fallbacks
Lide com recusas de forma elegante e distinga-as de bloqueios de classificadores.
Server-Side Fallbacks & Fallback Credit
The Opus 5 beta that turns a Fable 5 / Opus 5 refusal into a normal answer inside one API call — plus the credit token that stops you from paying the prompt cache twice on the retry.
Erros, Limites de Taxa e Confiabilidade
O mapa de erros, retentativas com backoff, níveis de limite de taxa e migração de modelos.
Memória e Edição de Contexto
Dê aos agentes memória persistente e contexto auto-podado — a ferramenta memory do lado do cliente mais a edição de contexto que limpa resultados de ferramentas obsoletos para sobreviver a execuções longas.
Programmatic Tool Calling
Let Claude call your tools from Python inside the code-execution sandbox — cutting round-trips and keeping intermediate results out of context.
The Admin API: automate your Claude org
The operator's guide to Anthropic's Admin API — list members, invite users, run RBAC groups, offboard cleanly. Console + Claude Enterprise, with the seat-pool and SSO gotchas that trip real teams.
Inference Hooks: inline DLP for Claude Enterprise
The operator's guide to Claude Enterprise Inference Hooks — the HTTPS webhook that lets your DLP server allow or deny every prompt across chat, Claude Code, and Cowork, before the model sees it. Full spec, verdict schema, signing, rollout playbook.