03 · The 2026 core skill

 Context engineering .

 Prompting is what you say; context engineering is everything the model can see. As agents got longer-running, deciding what fills the context window — instructions, retrieved documents, examples, tool outputs, history — and in what order became its own discipline (Mei et al., 2025; Anthropic, 2025).

 The constraint

 The context window is a budget

 A token is roughly ¾ of an English word. The context window is the fixed number of tokens the model can attend to per call — and everything competes for it: the system prompt, conversation history, retrieved data, the current question, and room for the answer itself.

 HF Agents Course tie-in (this week’s reading): context is the agent’s working memory . And like any working memory: more is not better.

 Context rot As the window fills with long or noisy input, accuracy and instruction-following degrade : models “lose the middle,” forget rules stated early, and drift — documented across 2024–25 long-context evaluations, including Chroma’s Context Rot report (2025).
 Practical rule: put the most important instructions first and last, and test at the lengths you’ll actually run.

 Curation over accumulation: include what’s relevant, exclude the rest — top-k retrieval instead of the whole manual (RAG preview, Week 7), summarize old history, drop stale turns. The anti-pattern: dumping everything into every call “just in case.”

 Interactive · Exercise B preview

 The packing game: fit the agent in 4,000 tokens

 You’re shipping a customer-support agent. Everything below “wants” to be in context — but it totals far more than your 4,000-token budget (you must leave ≥500 tokens of room for the answer). For each item choose Keep · Summarize · Drop , then check your packing.

 News hook · 2026

 Even Anthropic cut 80% of Claude Code’s system prompt

 Anthropic engineer Thariq Shihipar reported removing over 80% of Claude Code’s system prompt for Claude 5–generation models — with no measurable loss on coding evals (Cat Wu & Thariq Shihipar, Claude Code team, interviewed at the AI Engineer World’s Fair, July 2026). What got cut: examples (“it was just more creative than the examples we gave it”), “do not do this” rules, and absolute instructions that were only 90% true.

 Read it carefully: this is a capability-driven simplification — “the model got better, so the prompt could get shorter.” It is NOT “prompts don’t matter.” And it’s Claude Code’s prompt (the CLI coding agent), not the Claude.ai app — older models kept the full prompt. Viral restatements over-generalize; cite the primary interview.

 The interview write-up (Simon Willison, Jul 2026) ↗  ·  News coverage (The Decoder) ↗ 

 >80%
 of the system prompt removed, no measurable eval loss — as reported, 2026.

 Three takeaways for your prompts: 
1 · Less can beat more.
2 · Match scaffolding to capability — don’t over-constrain a strong model.
3 · Curate, don’t stuff — context rot is real.

 The last fork

 Prompting vs fine-tuning

 Prompt / context engineering Fine-tuning 

 What changes What the model sees The model’s weights 

 Cost & speed Free · instant Data + compute + upkeep · weeks 

 Data needed None (a few examples) Hundreds+ curated examples 

 Best for Almost everything — start here Consistent style at massive scale · cutting prompt cost · deep domain behavior 

 The order: prompt → few-shot → better context / RAG → fine-tune, last . Every rung you skip is money and weeks you didn’t need to spend.