Fuentes
Las afirmaciones clave de este curso se apoyan en las fuentes de abajo. Las lecciones las citan con [^Sn]; los extractos exactos están bajo cada entrada. Las citas se mantienen en su inglés original.
S1 — Effective context engineering for AI agents — Anthropic Engineering
URL: https://www.anthropic.com/engineering/effective-context-engineering-for-ai-agents
- authority: authoritative-guide
El artículo extenso y autorizado de Anthropic Engineering sobre ingeniería de contexto, y la base central de este curso: la relación entre la ingeniería de contexto y la ingeniería de prompts, el presupuesto de atención y la degradación de contexto (context rot), la «altitud» de un prompt del sistema, las herramientas y los ejemplos como contexto, la recuperación justo a tiempo y el descubrimiento progresivo, la compactación, la toma de notas estructurada y la arquitectura de subagentes; casi todos los temas de este curso salen de aquí.
Cita clave:
"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
El artículo autorizado de Anthropic sobre patrones de agentes. Este curso lo usa para la definición de un agente como bucle, los costos y los errores acumulativos de la autonomía, y la regla de proporción: añadir complejidad solo cuando mejora los resultados de forma demostrable.
Cita clave:
"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
La retrospectiva de ingeniería de Anthropic sobre su sistema multiagente de investigación, con datos de primera mano sobre el aislamiento de contexto de los subagentes y la economía del contexto: los subagentes en paralelo mantienen cada uno una ventana de contexto independiente, los subagentes actúan como «filtros inteligentes» que comprimen los hallazgos para el agente principal, y aparecen los multiplicadores de tokens de los agentes y de los sistemas multiagente.
Cita clave:
"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
La página oficial de buenas prácticas de Claude Code (anthropic.com/engineering/claude-code-best-practices ahora redirige aquí con un 308). Muestra qué aspecto tiene la gestión de contexto en un producto real: la disciplina en torno a CLAUDE.md, /clear y la compactación automática, las investigaciones con subagentes que aíslan el contexto principal, y la regla de cabecera de que el contexto es el recurso más importante que hay que gestionar.
Cita clave:
"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
Un roadmap comunitario de código abierto organizado en torno a la ingeniería del arnés (escrito con ayuda de IA). Este curso toma solo sus afirmaciones de encuadre: la gestión de contexto como uno de los componentes del arnés, y su definición en una línea de la ingeniería de contexto. Nota: sus umbrales porcentuales, cifras de tokens, puntuaciones de benchmark, salarios y demás datos concretos nunca se citan como hechos; «Prompt engineering is dead as a standalone skill in 2026» es una afirmación de opinión de ese roadmap, así que cualquier cita tiene que atribuírsela como tal, y el cuerpo del curso prefiere el encuadre más prudente de S1, el de la «progresión natural».
Cita clave:
"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."