Why Model APIs Lack Dedicated Few-Shot Fields

General model APIs treat few-shot demonstrations as raw text or chat messages rather than structured, versionable task artifacts. While specialized models like iAmTime and LoRA Recycle train directly on explicit demonstration structures, frontier LLM platforms rely on unstructured prompts, XML tags, or massive context windows. An examination of in-context learning shows why adding an examples parameter requires clear boundaries, ordering guarantees, and separate execution semantics beyond simple token pasting. Key sources: • OpenAI, fine-tuning guide (wind-down notice) — https://developers.openai.com/api/docs/guides/fine-tuning • OpenAI, SFT/distillation guide, Jan 6 2027 cutoff — https://developers.openai.com/api/docs/guides/distillation • OpenAI deprecations page, entries dated 2026-10-14, 2026-10-01, 2026-06-03 — https://developers.openai.com/api/docs/deprecations Episode 6310 — myweirdprompts.com/n/6310 (permanent short link: myweirdprompts.com/236875)