Event 139: conclusion-shape trigger + v2 landing rebuild#95
Merged
Conversation
Event 139 — the engine's primary decision shape gets its live trigger, and the site teaches the v2 engine visually: - conclusion_guard.py (UserPromptSubmit): conservative EN+KR decision-question lexicon -> per-prompt marker + one factual context line; conclusion_gate.py (Stop): one-shot nudge, flag-before-block (livelock-proof), stop verdicts count as interrogated. Registered in plugin hooks.json + adapters. 15 new tests; live-smoked. - E5 leg: every consumed interrogation verdict enqueues sample-all into the Layer 8 spot-check queue (artifact-hash idempotent). - web/: five-beat one-scroll arc (hero -> engine map -> three layers -> proof -> practice -> install) on the editorial-light language; animated EngineFlow centerpiece (motion@12, IntersectionObserver pause, reduced-motion static map, lazy chunk); InterrogationStrip hero visual; ThreeLayers (beat 02); ProofSection (E1 story + MIRROR bars + protocol No.1 transcribed); InstallSection with the real first-interrogation transcript. Adversarial-review must-fixes applied: missing motion CSS + global reduced-motion end-states, fixture-data-labeled-live honesty fix, real-vs-fictional lesson-hash separation, v1 OG/search metadata, WCAG contrast floors, contiguous section numbering. Stale v1 sections deleted. - i18n: ko/es/zh ledes at v2 parity; DESIGN_V2_0 trigger section marked shipped. pnpm build green (12 pages) + lint clean; pytest 1313 + 54 subtests.
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What this is
The v2.0 Epistemic Engine's primary decision shape — a load-bearing conclusion handed to the operator — gets its live trigger, and epistemekernel.com is rebuilt to teach the engine visually in one scroll.
Kernel (conclusion trigger + E5)
conclusion_guard.py(UserPromptSubmit): conservative positive-system lexicon of decision-question shapes (EN + KR) → per-prompt marker + one factual context line. Ungoverned projects and misses are silent.conclusion_gate.py(Stop): one-shot nudge when a conclusion-class turn ends without a fresh verdict — flag written before the block (livelock-proof by construction); astopverdict counts as interrogated;stop_hook_activeshort-circuits.hooks.json, adapters, and operator settings. 15 new tests; suite 1313 + 54 subtests green. Live-smoked end-to-end.Website (five-beat one-scroll arc, editorial-light preserved)
Hero ("Sounding right isn't being right.") with the InterrogationStrip stamp sequence → 01 how it works: the animated EngineFlow centerpiece (decision → tiered claims → sealed fresh-context verifier → evidence stamps → opposition → verdict gate → lesson into chain; motion@12, off-screen pause, reduced-motion static map, lazy chunk) → 02 the three layers (Cognition/Structure/Memory + where it runs) → 03 proof (the 49-days/zero-protocols E1 story, MIRROR 0.60→0.14 bars, protocol №1 transcribed) → 04 the practice → 05 install (real commands + the real first-interrogation transcript).
Adversarial-review must-fixes all applied: the missing motion CSS + a global reduced-motion block with explicit end-states; fixture-data-labeled-"live" honesty fix in ProofSection; real-vs-fictional lesson-hash separation; v1 OG/search metadata rewritten; WCAG contrast floors (whisper→muted/ash, 9px→11px); contiguous section numbering via the new ThreeLayers; stale v1 sections deleted. ko/es/zh ledes at v2 parity.
Verification
pnpm buildgreen (12 pages) + lint clean · pytest 1313 + 54 subtests · playwright full-page + hero screenshots inspected · conclusion hooks live-smoked under registered interpreter.Note: branch includes
chore(chkpt)auto-commits from the session-limit recovery; content is identical to the reviewed tip.