Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 11 additions & 0 deletions .okf/build/test-gates.md
Original file line number Diff line number Diff line change
Expand Up @@ -703,3 +703,14 @@ That mismatch cost two confident wrong rejections of a correct review finding on
#538 - §5.2 looked up in the v0.1 copy, twice - and a §9/§11 slip, since v0.1 §9
was conformance while v0.2 §9 is log structure. When a spec section is cited,
resolve WHICH copy before disputing it.

## Verifying production builds locally

A plain production build hard-codes absolute asset URLs at the live domain, so
a browser pointed at the local tree **silently loads the LIVE site's
stylesheet** - every computed-style check then measures the wrong CSS and
passes or fails meaninglessly. Build with `BASE_URL` set to the local origin
and confirm the page loads its own fingerprinted `.min.<hash>.css` before
trusting any computed value (caught 2026-08-21 during the pilot column-fix
verification; the earlier "inconclusive" production check had exactly this
cause).
16 changes: 16 additions & 0 deletions .okf/log.md
Original file line number Diff line number Diff line change
Expand Up @@ -51,6 +51,22 @@ make it green: restructure same-day entries under one heading, and add
`timestamp` to the 23 concepts missing it (anchored to each file's last commit
time, which is verifiable - never invented).

## 2026-08-21 - three register pilots shipped; the comparison is controlled; Editorial recommended

All three register pilots merged on the /next/ rail (#561 rescue-room, #562
editorial, #563 precision, #564 cross-pilot column fix): one landing anatomy,
three skins, copy parity verified at DOM level (identical text tokens, tags,
class sequences). Four-persona panel + fable critique recorded in 20.08:
pilot vote B 2 / C 1 / A 1; artifact head-to-head Editorial 4-1; recommendation
= Artifact Editorial + graft package (Precision's ledger table for data,
Rescue Room's mobile rhythm patterns - which are ANATOMY patterns, portable;
its template-silhouette weakness is intrinsic). Paul decides on the live URLs.
Durable instrument lesson added to test-gates: local production checks must
set BASE_URL or the browser silently loads the LIVE site's CSS. The shared
mobile table-clip defect (all three pilots, column 2 off-edge without
affordance) is deliberately deferred to the winner's graft round - fixing it
three times buys nothing.

## 2026-08-21 - register pilot B, and the two ways a second pilot breaks the first pilot's gates

`/next/pilots/editorial/fractional-cto/` ships the "Artifact Editorial" register
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -152,10 +152,12 @@ same commit · browser-measured AA + seams on anything touching a rendered page.
| Track A: judging board | VOTE RUN — **D1 wins 4–1** (20.07) — then SUPERSEDED by the register-pilots flow (§Flow revision): no early crown, compare pilots instead | — |
| Register research + shortlist | DONE — `10-19-research/10.01`: anatomy settled, 3 register candidates (A Rescue Room / B Artifact Editorial / C Engineered Precision), claims gate listed | — |
| Paul's Claude Design artifacts | A finished (`3dc819cf`, full landing page — the anatomy blueprint); B in progress (`8eb98dd6`) | pilots port from these; B refreshes when Paul finishes |
| Pilot A (Rescue Room) | NOT STARTED | port CD artifact `3dc819cf` to `/next/pilots/rescue-room/fractional-cto/`; claims gate FIRST (pricing/audit-promise/stay-stats need Paul) |
| Pilot B (Artifact Editorial) | NOT STARTED — existing `/next/services/fractional-cto/` has the skin but article anatomy | rebuild on the shared landing skeleton |
| Pilot C (Engineered Precision) | NOT STARTED | tokens from 10.01 research; same skeleton |
| Lightning demo | BLOCKED on pilots | side-by-side links + screenshots + 3-line verdicts per lens; discussion; Paul picks |
| Pilot A (Rescue Room) | **MERGED** #561 (`fd303af8c`) + column fix #564; canon-substituted copy | — |
| Pilot B (Artifact Editorial) | **MERGED** #562 (`451770a9e`) + column fix #564; fonts self-hosted, fake-bold fixed | — |
| Pilot C (Engineered Precision) | **MERGED** #563 (`45780ad6f`); DOM-verified copy parity across all three | — |
| Lightning demo | **BOARD LIVE** — `20.08`: pilot vote B 2/C 1/A 1; artifact head-to-head Editorial 4–1; fable rec = Editorial + graft package | **PAUL PICKS** on the live /next/ URLs |
| Visual-gate repair | #560 merged (`e1882483c`): tolerance 0.0001 (was 50x blind), shield documented, fossilization + viewport-only + perceptual-floor documented; machine exonerated (reds = #540-stale baselines) | follow-ups: re-record ~36 macOS + 22-key Linux dispatch; below-fold coverage |
| Homepage build (Editorial) | Blueprint ready (CD `8eb98dd6` Homepage.dc.html, team section removed) | GATED on Paul: audit-exhibit sourcing (real/SAMPLE/rebuild) + "code audit" offer naming |
| PR-1: v2 rail | **MERGED** #555 (`703fee12c`) — rail live: `/next/` noindexed staging, qtest holes closed (root layouts + v2 registry-derived keys), ADR-0006, mermaid source-aware. Screenshot job red = the documented 16-key #540 handoff only, verified key-by-key. | — |
| PR-2: styled pilot | **MERGED** #557 (`38f744805`) — full content ported (overview/faqs/tenure-derived stats), styled per D1, canon-exact Clutch proof line, browser-verified AA (button 5.13:1), −85.6% gzip CSS vs legacy page. Screenshot red = the same 16-key #540 handoff set, verified key-by-key. | compare links: `/next/services/fractional-cto/` vs `/services/fractional-cto/` |
| PR-3: flip | BLOCKED on lightning-demo winner + Paul's live review | flip = `type: next` on the real page + delete stub; checklist in ADR-0006 §5 |
Expand Down
Original file line number Diff line number Diff line change
@@ -0,0 +1,80 @@
# 20.08 — Register lightning-demo board (2026-08-21)

**Decision owner: Paul.** This board carries the pilots, the panel's votes, and
the fable recommendation. Nothing here is decided until Paul says so.

## The three pilots — one anatomy, three skins

Identical copy on all three, verified at DOM level (553 text tokens, one md5;
identical tag and class sequences). The only variable is the register.

| | Live (noindexed staging on production) | Local |
|---|---|---|
| A — Rescue Room (warm humanist) | https://jetthoughts.com/next/pilots/rescue-room/fractional-cto/ | `http://localhost:<port>/next/pilots/rescue-room/fractional-cto/` |
| B — Artifact Editorial (serif editorial) | https://jetthoughts.com/next/pilots/editorial/fractional-cto/ | …`/next/pilots/editorial/fractional-cto/` |
| C — Engineered Precision (ruled technical) | https://jetthoughts.com/next/pilots/precision/fractional-cto/ | …`/next/pilots/precision/fractional-cto/` |

Legacy comparison: https://jetthoughts.com/services/fractional-cto/

## Panel vote (4 personas, sonnet; distinct lenses; forced self-objection)

| Voter | 1st | 2nd | Core argument |
|---|---|---|---|
| Sam — burned founder (trust in 5s) | **B** | C | "reads like a document I can trust, not a pitch I'm being sold" |
| CFO — buyer (price justification) | **C** | B | "the itemized ledger that keeps this firm from ever surprise-billing me" |
| Mobile UX (390px rhythm) | **A** | C | threaded timeline, one-line nav, CTA color constancy start-to-finish |
| Brand strategist (differentiation + scale) | **B** | C | B *performs* "we show our work"; A is the template silhouette to escape |

First choices **B 2 · C 1 · A 1**. C sits in every voter's top two (the safe
choice); **A took zero trust/brand/buyer votes** — warm-rounded consistently
read as template. Strongest self-objections, recorded: B "editorial polish is
what a firm hides behind when substance is missing" (Sam) and "could read as a
media property, not engineers" (Brand); C "am I paying for software or a
person's judgment?" (CFO); A "softer/consumer-app, less rigorous" (Mobile).

## Artifact-level head-to-head (Paul's review request, 2026-08-21 evening)

Paul asked for a review+vote between the two Claude Design artifact sets:
the **Artifact Editorial set** (`8eb98dd6`: homepage + service page — the only
entry proven as a SYSTEM across two pages) vs **Rescue Room** (`3dc819cf`:
service page only). Restricting each recorded ballot to those two:
**Editorial 3 — Rescue Room 1** (Sam, CFO, Brand → Editorial; Mobile → Rescue
Room), plus the fable vote → **Editorial 4–1**. Rescue Room's durable
contributions are its ANATOMY (the shared skeleton) and its mobile rhythm
patterns — both survive regardless via the graft package.

## Fable recommendation: **B — Artifact Editorial, with a graft package**

The tie-breaking observation: the mobile voter's pro-A arguments are **anatomy
patterns, not register properties** — a threaded timeline, one-line nav, and
CTA color constancy graft into any skin. A's weakness (template silhouette) is
intrinsic; B's weaknesses are all fixable inside the register. And B is the
only register that structurally matches the content strategy Paul's own Claude
Design artifacts converged on — selling with documents (redacted audit
exhibits, Friday-report cards) demands an editorial system.

Graft package (applies to B if chosen):
1. **C's ruled-ledger table treatment** for data surfaces (the CFO vote, kept).
2. **A's mobile rhythm patterns**: threaded timeline, one-line nav at 390,
CTA stays accent-filled through the footer (B currently breaks this).
3. B mobile fixes: reduce all-caps density at small sizes; thread the step
boxes.

## Defects on the board (status)

- **A+B muted-column**: FIXED — #564 merged (`0c23219e8`); both pilots'
comparison columns now compute their muted tokens (A 5.95:1, B 5.74:1).
- **Shared mobile table clip** (all three clip column 2 off-edge, no scroll
affordance): affects all three equally → does not distort the comparison;
fixed in the winner's graft round, not three times.
- First-fold-only screenshot coverage (below-fold unverified by the suite;
browser-verified instead) — tracked follow-up.

## Decision protocol

Paul reviews the live links (phone + desktop), reads the votes, then picks:
**A / B / C / none (rerun with new directions)**. The two losing pilots'
stubs+CSS are deleted; the winner's register gets the graft round, then the
service-page flip (`type: next` on the real page) per ADR-0006 §5, then the
homepage build from Paul's Claude Design artifact (claims question pending:
the audit exhibit needs a real source or an honest SAMPLE label).