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 — Effective context engineering for AI agents — Anthropic Engineering
URL: https://www.anthropic.com/engineering/effective-context-engineering-for-ai-agents
- authority: authoritative-guide
O artigo longo e autoritativo da engenharia da Anthropic sobre context engineering, e a base central deste curso: a relação entre context engineering e prompt engineering, o orçamento de atenção e o apodrecimento do contexto, a “altitude” de um system prompt, ferramentas e exemplos como contexto, a recuperação just-in-time e a descoberta progressiva, a compactação, as notas estruturadas e a arquitetura de subagentes — quase todo tópico deste curso vem dele.
Citação principal:
"we view context engineering as the natural progression of prompt engineering" "Prompt engineering refers to methods for writing and organizing LLM instructions for optimal outcomes" "the set of strategies for curating and maintaining the optimal set of tokens (information) during LLM inference" "An agent running in a loop generates more and more data that could be relevant for the next turn of inference" "LLMs have an "attention budget" that they draw on when parsing large volumes of context" "Every new token introduced depletes this budget by some amount" "as the number of tokens in the context window increases, the model's ability to accurately recall information from that context decreases" "some models exhibit more gentle degradation than others, this characteristic emerges across all models" "These factors create a performance gradient rather than a hard cliff" "context, therefore, must be treated as a finite resource with diminishing marginal returns" "agents that operate over multiple turns of inference and longer time horizons" "engineers hardcoding complex, brittle logic in their prompts to elicit exact agentic behavior" "vague, high-level guidance that fails to give the LLM concrete signals for desired outputs" "specific enough to guide behavior effectively, yet flexible enough to provide the model with strong heuristics" "tools should be self-contained, robust to error, and extremely clear with respect to their intended use" "building tools that are well understood by LLMs and have minimal overlap in functionality" "returning information that is token efficient and by encouraging efficient agent behaviors" "stuff a laundry list of edge cases into a prompt" "curate a set of diverse, canonical examples that effectively portray the expected behavior of the agent" "maintain lightweight identifiers (file paths, stored queries, web links, etc.)" "the metadata of these references provides a mechanism to efficiently refine behavior" "allows agents to incrementally discover relevant context through exploration" "retrieving some data up front for speed, and pursuing further autonomous exploration at its discretion" "CLAUDE.md files are naively dropped into context up front, while primitives like glob and grep" "taking a conversation nearing the context window limit, summarizing its contents, and reinitiating a new context window" "passing the message history to the model to summarize and compress the most critical details" "preserves architectural decisions, unresolved bugs, and implementation details while discarding redundant tool outputs" "the agent regularly writes notes persisted to memory outside of the context window" "Like Claude Code creating a to-do list, or your custom agent maintaining a NOTES.md file" "Claude playing Pokémon demonstrates how memory transforms agent capabilities in non-coding domains" "specialized sub-agents can handle focused tasks with clean context windows" "returns only a condensed, distilled summary of its work (often 1,000-2,000 tokens)" "the detailed search context remains isolated within sub-agents" "compaction, structured note-taking, and multi-agent architectures" "maintain coherence, context, and goal-directed behavior over sequences of actions"
S2 — 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 a definição de agente como um loop, os custos e os erros que se acumulam com a autonomia, e a regra de proporção — acrescentar complexidade só quando isso melhora os resultados de forma demonstrável.
Citação principal:
"They are typically just LLMs using tools based on environmental feedback in a loop." "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." "you should consider adding complexity only when it demonstrably improves outcomes."
S3 — 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, com dados de primeira mão sobre o isolamento de contexto de subagentes e a economia de tokens: subagentes em paralelo mantêm cada um uma janela de contexto independente, os subagentes agem como “filtros inteligentes” que comprimem os achados para o agente líder, e os multiplicadores de tokens de agentes e de sistemas multiagentes.
Citação principal:
"Subagents facilitate compression by operating in parallel with their own context windows" "distributes work across agents with separate context windows to add more capacity for parallel reasoning" "agents typically use about 4× more tokens than chat interactions" "multi-agent systems use about 15× more tokens than chats" "Multi-agent systems work mainly because they help spend enough tokens to solve the problem." "token usage by itself explains 80% of the variance, with the number of tool calls and the model choice" "condensing the most important tokens for the lead research agent" "The essence of search is compression: distilling insights from a vast corpus." "the subagents act as intelligent filters" "agents summarize completed work phases and store essential information in external memory" "agents can spawn fresh subagents with clean contexts while maintaining continuity through careful handoffs"
S4 — Best practices for Claude Code — Claude Code Docs
URL: https://code.claude.com/docs/en/best-practices
- authority: official-docs
A página oficial de boas práticas do Claude Code (anthropic.com/engineering/claude-code-best-practices hoje redireciona para cá com um 308). Mostra como a gestão de contexto se parece em um produto real: a disciplina em torno do CLAUDE.md, o /clear e a compactação automática, as investigações por subagente isolando o contexto principal, e a regra-título de que o contexto é o recurso mais importante a administrar.
Citação principal:
"Claude's context window fills up fast, and performance degrades as it fills." "The context window is the most important resource to manage." "This matters since LLM performance degrades as context fills." "CLAUDE.md is a special file that Claude reads at the start of every conversation." "CLAUDE.md is loaded every session, so only include things that apply broadly." "Claude loads them on demand without bloating every conversation." "Keep it concise. For each line, ask: "Would removing this cause Claude to make mistakes?" If not, cut it." "Bloated CLAUDE.md files cause Claude to ignore your actual instructions!" "If your CLAUDE.md is too long, Claude ignores half of it because important rules get lost in the noise." "reset context between unrelated tasks. Long sessions with irrelevant context can reduce performance." "Run /clear between unrelated tasks to reset the context window entirely" "During long sessions, Claude's context window can fill with irrelevant conversation, file contents, and commands." "Claude Code automatically compacts conversation history when you approach context limits, which preserves important code and decisions while freeing space." "Since context is your fundamental constraint, subagents are one of the most powerful tools available." "Subagents run in separate context windows and report back summaries" "Scope investigations narrowly or use subagents so the exploration doesn't consume your main context." "A clean session with a better prompt almost always outperforms a long session with accumulated corrections."
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 gestão de contexto como um dos componentes do harness, e sua definição em uma linha de context engineering. Observação: seus limiares percentuais, números de tokens, notas de benchmark, valores de salário e outros dados específicos nunca são citados como fato; “Prompt engineering is dead as a standalone skill in 2026” é uma asserção opinativa desse roadmap — qualquer citação precisa atribuí-la como tal, e o corpo do curso prefere o enquadramento mais cuidadoso de “natural progression” da S1.
Citação principal:
"Same model, different harness, completely different result." "the harness is the union of:" "context engineering: deciding what tokens are in front of the model at every step of the loop" "Prompt engineering is dead as a standalone skill in 2026."