Paul BakausandClaude Fable 5 a38a0765a9 skill v4.0.0-alpha.1: always-loaded core — brief-wins, subject grounding, token/self-check process, inline craft floor + registers ...

One-shot evals on Fable 5 (impeccable-evals notes/fable-oneshot-craft-plan.md)
showed the reference-file architecture failing: models skip the register
reads, so most design guidance never reaches them, and skill-on collapses
toward bare-model output (0/9 pairwise wins vs frontend-design on r10).

SKILL.md is now self-contained for one-shot work: persona, the-brief-wins
rule, ground-it-in-the-subject, a plan/tokens/signature/self-check process
gate, commitment guidance, a compact inline craft floor, and distilled
brand/product registers. Reference files remain as sub-command flows and
optional depth. The enumerated absolute-bans list is retired from prose;
mechanical slop enforcement moves to the detector/hook.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

2026-07-10 10:46:10 -07:00

..

agents

Live mode: staged AI copy edits (#158)

2026-05-28 19:02:12 -07:00

reference

Expand init to capture positioning, conversion, and proof context (#315)

2026-07-09 16:20:20 -07:00

scripts

Add OpenAI plugin submission bundle (#363)

2026-07-09 17:09:13 -07:00

SKILL.src.md

skill v4.0.0-alpha.1: always-loaded core — brief-wins, subject grounding, token/self-check process, inline craft floor + registers

2026-07-10 10:46:10 -07:00