Remediation of the verified findings from the 2026-06-12 plan-set audit
(V1_V7_PLAN_SET_AUDIT_2026-06-12.md §4, §5, §8) against the V1 planning docs.
Every change below was made as an individual judged edit after reading the cited
lines. No checkbox was changed to [x]; re-marks are [x]→[~] only, each with
an inline dated reason on the task line.
1. Substrate count and Aje glossary (V1/features.md)#
- features.md:23–25 — "Five platform substrates — Sophia, Iris, Psyche,
Lilith, and Isis" → "Six platform substrates — Sophia, Iris, Psyche, Lilith,
Isis, and Aje", adding "non-custodial payment settlement" to the
responsibility list. Reason:
V1/ARCHITECTURE.md(Subsystem Glossary and intro) counts six platform substrates including Aje, promoted 2026-05-11; features.md had never been updated. - features.md Subsystem Glossary ("Platform substrates") — added an Aje
entry mirroring
ARCHITECTURE.md:167: library-only blockchain/Web3 payment substrate (Yoruba orisha of wealth), non-custodial settlement, merchant invoices, per-chain providers, wallet management; cross-links to the existing features Crypto Payments section (~features.md:5417) andARCHITECTURE.md§ Aje. Reason: glossary listed five substrates and never defined Aje despite full Aje content in the payments section. - ARCHITECTURE.md:134 (V1 Product Promise) — "Five platform substrates" → "Six platform substrates … and non-custodial payment settlement (Aje)". Reason: internal contradiction with ARCHITECTURE.md:24 and :169–171 ("six platform substrates in total").
2. Living Scenes template-name drift (V1/ARCHITECTURE.md)#
- ARCHITECTURE.md:1296–1297 — Nyx "Sky Awe" → "Sky Briefing"; Metis "Lesson Arc" → "Lesson Visualizer". Reason: features.md:3631–3633/3956/3964 and TODOS.md §25 uniformly use "Sky Briefings" / "Lesson Visualizers".
- ARCHITECTURE.md:1331 (Conductor-flow mermaid) — "Nyx Sky Awe" → "Nyx Sky Briefing". Same reason.
- ARCHITECTURE.md:~1573 (Generation Audience Tiers,
Customerrow) — surfaces extended from "Living Offerings (Arete), Contemplative Arcs (Tara), explainers (Veritas)" to the full five customer-tier generation surfaces, adding "Sky Briefings (Nyx)" and "Lesson Visualizers (Metis)" and naming the Veritas surface "Grounded Explainers". Reason: features.md:3220–3224 lists five customer-tier surfaces; ARCHITECTURE listed three.
3. Bellona / Hathor / Neith (Studio-only substrates)#
Verified first: features.md:82–97 declares them in-V1 Studio substrates
(confirmed in-v1 2026-05-29); inventory exists at
WALKTHROUGH/studio/domain-bridges/{bellona,hathor,neith}/ (82 / 74 / 1 docs);
Studio routes exist at apps/oshun/web/src/app/studio/{bellona,hathor,neith}/;
implementations at libs/bellona/* and libs/hathor/*. They had zero mentions
in ARCHITECTURE, TODOS, and DEPENDENCIES.
- ARCHITECTURE.md:24–26 — "three cross-domain support subsystems" → "six cross-domain support subsystems (Aja, Yemaya, and Themis, plus the Studio-only Bellona, Hathor, and Neith)".
- ARCHITECTURE.md Subsystem Glossary ("Cross-domain support" table) — added Bellona, Hathor, and Neith rows.
- ARCHITECTURE.md § Cross-Domain Support (body table, ~:927) — added
Bellona, Hathor, and Neith rows (adapter paths + V1 roles per the features
glossary) and a paragraph noting they are Studio-only V1 substrates with
inventory at
WALKTHROUGH/studio/domain-bridges/and backlog at TODOS § 34. - DEPENDENCIES.md § 2 ("Cross-domain support" table) — added Bellona,
Hathor, and Neith registry rows pointing at
libs/bellona/*,libs/hathor/*, and the Neith Studio workspace. - TODOS.md — added new section "34. Studio Substrate Bridges (Bellona,
Hathor, Neith)" (plus TOC entry) with four honest
[ ]verification tasks pointing at the bridge inventory, routes, and libs. Nothing marked[x].
4. Stale provenance#
- TODOS.md:3 — dropped the stale "(3,004 lines)" features.md line count (features.md is now ~6,242 lines); noted the removal inline.
- features.md:4 — "Source:
V1/TODOS.md(with_P2companion)" → "Source:V1/TODOS.md". Reason: no_P2file exists inV1/(verified by directory listing). - features.md:15 — removed "(and its
_P2companion)" from the purpose paragraph. Same reason.
5. Launch-locale enumeration (V1/features.md)#
- features.md:~2670 (first "V1 launch locale" reference, Lilith crisis
localization bullet) — enumerated the launch-locale set once: en-US,
es-US, fr-FR, de-DE, ar, he, ja-JP, pt-BR (en-US default; ar/he RTL), citing
where the set is configured. Repo evidence:
OSHUN_LAUNCH_LOCALESinlibs/oshun/i18n/src/index.ts(canonical catalog, language-preference mapping, RTL set, fallback chain<requested>→<language-only>→en-US), consumed byapps/oshun/web/src/i18n/config.ts. Note:apps/oshun/mobile/src/i18n/locale.tsresolves device locales against a broader 25-localeSUPPORTED_LOCALESlist, but the canonical launch catalog is the 8-locale@oshun/i18nset; the features.md sentence cites the canonical source.
6. Surface table (V1/ARCHITECTURE.md:183–193)#
- Added rows for Oshun Telegram Bot (
apps/oshun/telegram-bot/) and Oshun Telegram Mini App (apps/oshun/telegram-miniapp/), both citing TODOS § 26 (Messaging Channels). Reason: both apps exist in the repo (verifiedls apps/oshun/) and were missing from the table.
7. Completion-state banner (V1/TODOS.md)#
- Added a clearly-marked blockquote banner after the Purpose section (modeled on
V5/V5_TODOS.md:40–54): the 1,639[x]marks (verified count:grep -cE "^\s*- \[x\]"→ 1,639 at audit time) predate the root-level audit chain; the operative residual truth isV1_RESIDUAL_BACKLOG_2026-06-11.md(~250 findings incl. 12 P0-SEC),docs/audits/V1_RESIDUAL_AUDIT_2026-06-11/, andV1_PRODUCT_COHERENCE_AUDIT_2026-06-12.md(all verified present at repo root);[x]must not be treated as authoritative without consulting those.
8. Re-marked not-locally-actionable launch-readiness items (V1/TODOS.md)#
Each re-mark is [x]→[~] with an inline dated reason, per the doc's own
convention (TODOS.md "How to use this file", pre-edit lines 53–55). Individually
judged:
- §28 (:~4996) Pre-GA penetration testing + red-team exercises — external vendor
exercise (the doc's own
[~]example). - §33.1 Internal dogfood drill (≥ 30 days) — real-world drill.
- §33.1 Safety review pass — real-world review-board sign-off.
- §33.1 Quality review pass — human review event (eval scores repo-verifiable; the pass itself is not).
- §33.1 Private beta with measurable target metrics — real-world beta cohort.
- §33.1 Beta-blocking-issue remediation + signoff — depends on the beta.
- §33.2 General Availability approval captured — real-world approval.
- §33.2 Rollout plan approved — approval is a real-world decision.
- §33.2 Rollback plan approved — same.
- §33.2 Post-deploy monitoring dashboards live — requires production.
- §33.2 Canary analysis automation enabled — requires production rollout infrastructure.
Judged and left [x]: "Launch go/no-go checklist completed at
docs/launch/go-no-go.md" (the artifact exists on disk — locally verifiable);
§33.3 critical-journey verifications and §33.4 exit-criteria signoffs (covered
by the completion-state banner; not individually named by the audit). The App
Store / Play Store submission tasks at TODOS.md:~442–446 sit in § 2 (outside the
Launch Readiness section this fix was scoped to) and were left to the banner.
9. Dependency gap closure section (V1/TODOS.md)#
- Added new section "35. Dependency Gap Closure (from DEPENDENCIES
self-flagged launch gaps)" (plus TOC entry) with one unchecked
[ ]task per self-flagged gap, each citing the DEPENDENCIES section and post-edit line number: missing mobile Expo pins (deps § 6, :201–207), no C2PA mint/verify SDK (deps § 19, :546), undocumentedARETE/VERITAS/NYX/METIS_DATABASE_URL(deps § 21, :592–602), noNYX_*_URLentries (deps § 18, :521–524), no ElevenLabs SDK pin (deps § 9, :260).
10. Generated HTML renders (V1/features.html, V1/ARCHITECTURE.html)#
The two .html files are generated render snapshots of the markdown and
contained the same stale strings. The exact fixes from items 1, 2, and 4 were
mirrored into them (Sky Awe → Sky Briefing, Lesson Arc → Lesson Visualizer, five
→ six platform substrates, _P2 references removed) so no stale string survives
anywhere in V1/. Caveat: the HTML snapshots remain stale renders overall
(they do not include the new Aje glossary entry, Telegram surface rows,
cross-domain additions, or launch-locale enumeration) and should be regenerated
from the updated markdown by whatever pipeline produced them; full regeneration
was out of scope for this remediation.
Files changed#
| File | Change |
|---|---|
V1/features.md |
Six-substrate count; Aje glossary entry; _P2 references removed; launch-locale enumeration |
V1/ARCHITECTURE.md |
Six-substrate promise; six cross-domain subsystems; Bellona/Hathor/Neith glossary + body rows; template-name fixes; Customer-tier surfaces completed; Telegram surface rows |
V1/DEPENDENCIES.md |
Bellona/Hathor/Neith cross-domain registry rows |
V1/TODOS.md |
Provenance fix; completion-state banner; 11 individual [x]→[~] re-marks; new §34 and §35 + TOC entries |
V1/features.html |
Mirror-patch of stale strings (six substrates, _P2 removal) in the generated render; regeneration still needed |
V1/ARCHITECTURE.html |
Mirror-patch of stale strings (six substrates, Sky Briefing, Lesson Visualizer) in the generated render; regeneration still needed |
V1/REMEDIATION_2026-06-12.md |
This log (new) |