Eve library

Arete · One missed morning, five owners · receipt · 12 September 2026

Guide

Revision 1

On this page

Chapter one of The other rooms, delivered in the Eve edition on 12 September 2026. It merges the nine-slide arete introduction and the forty-three-slide arete-humane-engagement field guide into one chapter under the arete id, and opens the track. The brief is authoring/arete-chapter.md; the track brief is authoring/the-other-rooms-track.md. Machine-readable state is in state.json.

What was delivered#

53 slides in the guide: 51 authored here, plus lilith-v1-0 and room-maturity carried unchanged from the Meet Lilith track at their existing positions. Nine slides are new — a cover, six section dividers, a six-word glossary and a close. Four slides of the old introduction belong to the reviewed foundation (foundation-coverage.json) and are rewritten in place under their own identifiers; three more are absorbed by the deeper treatment of the same subject, and five humane slides join a neighbour. 62.1 minutes of narration.

Merged away, with the slide that now carries the teaching:

Slide Absorbed by
arete-boundary arete-humane-opening
arete-worked arete-humane-worked-trace
arete-failure arete-humane-canonical-gap
arete-humane-friction-domain arete-humane-friction-contract
arete-humane-rest-cadence arete-humane-miss-counting
arete-humane-concept-links arete-humane-cross-domain
arete-humane-persistence-map arete-humane-current-matrix
arete-humane-related-map arete-close

The first three were in the published field guide and are recorded in MERGED_AWAY_SLIDES; the other five were added after it, so that baseline never held them.

Six sections of five to eight slides. A section-divider names at most six sections, so a fifty-one-slide chapter cannot also hold every section to the usual seven; three of them run to eight.

Evidence#

  • Source harness. tools/presentations/tests/pinned-arete-humane-engagement.test.mjs re-run on this host: 24 cases pass, covering the route tree, the check-in contract and its refinements, the two policy vocabularies, the five recovery stages, the consumer write and its projection, the streak fold, the absent canonical records, the review models, the mobile completion, the friction taxonomies, the cross-domain builders and the Living Offering.
  • Four captures of actual routes at the pin under a development member session: the hub with its BFF deliberately unavailable, the recovery preview, the review preview, and the offering with text typed into it. Each carries its dimensions and SHA-256 in state.json.
  • An independent source check of the draft found thirty-one wrong, overstated or unsupported claims. Every one was re-verified against the pin and fixed before the first build. The largest corrections:
    • the recovery record has no triggerCount field; the rule is that a non-manual record must cite at least one piece of evidence;
    • the stored rows keep the exact posted status — what the route never accepts is a percentage or a reason, so engagement_percent is synthesised (100/50/0) and status_reason is written null while the member's note is kept as reflection;
    • the recovery engine does run on the member path (the hub, the streak read and the check-in write all return its stage); the /arete/recovery page is the one that does not call it;
    • one check-in writes two durable rows, and the canonical one is insert-once, so an amended day never reaches it;
    • the habit update writes three fields and the best streak is returned but not stored;
    • rest-days is not a canonical cadence — it exists only in the recovery engine's own input type, which maps it to the daily counter;
    • the two friction vocabularies share no value, so every mapping is a judgement, and a third, narrower vocabulary exists in libs/arete/habits;
    • the durable offering record carries no manifest or hashes — those come from a separate receipt route that stores nothing — and an empty intention skips the safety screen entirely;
    • four of the five cross-domain builders have app callers; the Tara resolver is the one that has none.

Checks#

Check Result
Build 1,538 slides, 662 sources; freshness check clean
Python suite 481 tests pass
Node suite 489 tests pass
Browser sweep 53 slides × 1440, 1280 with audio chrome, 390 and print — 0 findings
PDF 53 pages, exact authored visual text, both tables complete
Narration 1,538 tracks verified, nothing pending
Bindings check-center-coverage.py --check reports no stale bindings

Eleven layout findings were fixed before the sweep passed: every five-card grid overflows 1280×720, so nine of them became four-card grids; two case studies became three-node cases; and the maturity table lost its widest column, which the notes already carried.

What this chapter does not establish#

Nothing here touches a deployed service, a real member account, generated media, or delivery to another person. The captures are of actual routes rendered from the pinned revision with declared input — one with its service deliberately unavailable, which is evidence about the page, not about an outage. No claim is made that the mobile check-in reaches a route, or that any canonical missed-day, streak, recovery or plan record is written.