Fontes

As principais afirmações deste curso se apoiam nas fontes abaixo. As lições as citam com [^Sn]; os trechos exatos ficam sob cada entrada.

S1 — How we built our multi-agent research system — Anthropic Engineering

URL: https://www.anthropic.com/engineering/multi-agent-research-system

• authority: authoritative-guide

A retrospectiva de engenharia da Anthropic sobre seu sistema de pesquisa multiagente, e a base central deste curso: agentes têm estado e os erros se acumulam; não dá para recomeçar do zero — é preciso retomar do ponto em que o erro aconteceu; a confiabilidade vem de juntar a adaptabilidade do modelo a salvaguardas determinísticas, como lógica de retry e checkpoints regulares; e, na hora de fazer deploy, um agente em execução pode estar em qualquer ponto do seu processo.

Citação principal:

"Agents can run for long periods of time, maintaining state across many tool calls." "Agents are stateful and errors compound." "Without effective mitigations, minor system failures can be catastrophic for agents." "When errors occur, we can't just restart from the beginning: restarts are expensive and frustrating for users." "Instead, we built systems that can resume from where the agent was when the errors occurred." "This means we need to durably execute code and handle errors along the way." "the adaptability of AI agents built on Claude with deterministic safeguards like retry logic and regular checkpoints" "letting the agent know when a tool is failing and letting it adapt works surprisingly well" "we use rainbow deployments to avoid disrupting running agents, by gradually shifting traffic from old to new versions" "whenever we deploy updates, agents might be anywhere in their process." "agents summarize completed work phases and store essential information in external memory before proceeding to new tasks"

S2 — Checkpointing — Claude Code Docs

URL: https://code.claude.com/docs/en/checkpointing

• authority: official-docs

A documentação oficial do checkpointing do Claude Code, que mostra a forma completa e de nível de produto do checkpoint/rewind/fork: um checkpoint automático antes de cada prompt da pessoa usuária, salvo junto com a conversa, conversa e código restauráveis separadamente, mudanças feitas por bash não rastreadas, e nenhum substituto para o controle de versão. Este curso a usa como ponto de referência de nível de produto, não como a ferramenta que está sendo ensinada.

Citação principal:

"checkpointing automatically captures the state of your code before each user prompt." "Every user prompt creates a new checkpoint" "Claude Code saves checkpoints with the conversation, so you can still run /rewind after you resume a session" "Restore code and conversation: revert both code and conversation to that point" "Restore conversation: rewind to that message while keeping current code" "Restore code: revert file changes while keeping the conversation" "Exploring alternatives: try different implementation approaches without losing your starting point" "Recovering from mistakes: quickly undo changes that introduced bugs or broke functionality" "Checkpointing does not track files modified by bash commands." "Only direct file edits made through Claude's file editing tools are tracked." "Checkpoints are designed for quick, session-level recovery." "continue using version control, such as Git, for commits, branches, and long-term history." "To branch off and try a different approach while preserving the original session intact, use /branch or claude --continue --fork-session"

S3 — Building Effective AI Agents — Anthropic Engineering

URL: https://www.anthropic.com/engineering/building-effective-agents

• authority: authoritative-guide

O artigo autoritativo da Anthropic sobre padrões de agente. Este curso o usa para: agentes que pausam em checkpoints à espera de feedback humano, os custos mais altos e os erros que se acumulam com a autonomia, a confiança limitada ao longo de muitos turnos, e a regra de proporção — acrescentar complexidade só quando isso melhora os resultados de forma demonstrável.

Citação principal:

"Agents can then pause for human feedback at checkpoints or when encountering blockers." "The autonomous nature of agents means higher costs, and the potential for compounding errors." "The LLM will potentially operate for many turns, and you must have some level of trust in its decision-making." "it's also common to include stopping conditions (such as a maximum number of iterations) to maintain control." "you should consider adding complexity only when it demonstrably improves outcomes."

S4 — Handle tool calls — Claude API

URL: https://platform.claude.com/docs/en/agents-and-tools/tool-use/handle-tool-calls

• authority: official-docs

Dá a especificação dos campos e a regra de emparelhamento dos blocos de conteúdo tool_use e tool_result: todo tool_use precisa receber um tool_result correspondente, todos devolvidos juntos. A lição sobre retomada depende dessa regra para lidar com a chamada órfã que sobra quando uma queda acontece entre a execução da ferramenta e a gravação no livro-razão.

Citação principal:

"id: A unique identifier for this particular tool use block. ... name: The name of the tool being used. input: An object containing the input being passed to the tool, conforming to the tool's input_schema. ... tool_use_id: The id of the tool use request this is a result for. ... is_error (optional): Set to true if the tool execution resulted in an error." "Whichever strategy you use, return one tool_result for each tool_use block, all together in the next user message. Match each result to its call with tool_use_id, and put every tool_result block before any text content in that message."

S5 — The 2026 Agent Engineering Roadmap — GitHub (codejunkie99/agent-roadmap-2026)

URL: https://github.com/codejunkie99/agent-roadmap-2026

• authority: blog

Um roadmap comunitário de código aberto organizado em torno da engenharia de harness (escrito com auxílio de IA). Este curso toma dele apenas as afirmações de enquadramento: a persistência como um dos componentes do harness, e o trabalho dela — “checkpoint state every node so you can resume, rewind, fork.” Observação: seus limiares percentuais, números de tokens, notas de benchmark, valores de salário e outros dados específicos nunca são citados como fato.

Citação principal:

"the harness is the union of:" "persistence. Checkpoint state every node so you can resume, rewind, fork." "Same model, different harness, completely different result."