TT Claude Memory

project_session_2026_04_19_evening

Properties1
descriptionMonumental ~7-hour evening session producing the canonical brand-architecture.md v1.5 + frontmatter-conventions.md v1.0 + ideas/ folder + 10 CLAUDE.md embeddings + _archive_/ folder + retrofits across 4 canonical docs. Rich's 15-point decision document + subsequent directives drove the consolidation.

6 min read

Sunday 19 April 2026 evening, ~18:00-01:00. One of the most important strategy-consolidation sessions TT has had. Fresh sessions should read this to understand the scope of what changed.

Context entering the session

  • Earlier in the day: Option F v1.3 critical review (Ashley + Kimothi + Fowler books); Option G+ v1.1 approved (HNW + partner-matching pulled into Phase 1); Nova deprioritised; LL migration doc drafted; 24-month roadmap committed
  • Rich became concerned that competing documents + inconsistent filing had caused repeated rework
  • Specifically: two parallel brand-architecture docs (tt-brand-architecture.md v2.4 vs multi-brand-architecture.md v1.4) disagreed about whether InheritWills was a brand

What happened — phased narrative

Phase A — Discovery + problem-framing

  • Surveyed 225+ markdown files across 5 repos (testatetech/docs-strategy + code-mfi + code-ll + openinherit/code-standard etc.)
  • 82% of surveyed docs were status: draft — signal lost
  • 13 files in tt-products/ had zero frontmatter
  • Identified 12 collisions / orphans / staleness issues
  • Rich narrowed scope: leave openinherit/code-standard alone (website-build coupling); focus on TT brand clarity

Phase B — 15-point decision document

  • Rich returned with a Google Doc Prompts for Claude - save as pdf and drag in.md answering Claude’s 15-item brand-and-artefact inventory point-by-point
  • Then a second Google Doc (1).md with follow-up decisions on IW-first / InheritKit-next / Advisory BackOffice / will scanner / Catalogue position / LLv2 parallel
  • Every decision captured in docs-strategy/ideas/2026-04-19-15-point-decision-document.md as primary evidence (verbatim Rich quotes preserved)

Phase C — Canonical brand-architecture.md consolidation

  • Created docs-strategy/brand-architecture.md v1.0 absorbing all 15 decisions + closing directives
  • Superseded tt-brand-architecture.md v2.4 and multi-brand-architecture.md v1.4 (retained with supersession banners)
  • Iterated through v1.1 (build-order vs launch-order separated) → v1.2 (OpenInherit first in build order) → v1.3 (LL v1 removed from build chain + Independent Tracks §6.4) → v1.4 (OpenInherit public-release plan §6.5)
  • Sibling updates: scorecard v1.6 → v1.7 (status revert from approved to best-current-proposal-pending-product-requirements per Rich’s meta-critique); Option G+ v1.1 → v1.2 (Advisory BackOffice → InheritKit; will scanner → IW-only; Catalogue open; LLv2 parallel); roadmap v1.0 → v1.5 (5 reorders absorbing the directives)

Phase D — Filing-system conventions

  • Rich asked to expand frontmatter conventions so “gold is not lost” in tangents
  • Created docs-strategy/frontmatter-conventions.md v1.0 — 11-value canonical status vocabulary; gold-preservation fields (origin, related, open-questions, gold-value); ideas/ folder convention
  • Created docs-strategy/ideas/ with README + template
  • Retrofitted 5 of today’s tangents as ideas/YYYY-MM-DD-*.md files with full frontmatter

Phase E — “All roads point back” embedding

  • Added global ~/.claude/CLAUDE.md §17 — TT canonical-docs section so every Claude session worldwide sees the pointers
  • Created docs-strategy/CLAUDE.md (entry point for strategy repo)
  • Created code-inheritwills/CLAUDE.md (empty repo; IW is #1 launch priority)
  • Prepended “Canonical TT Strategy Context” sections to 7 existing CLAUDE.md files: code-mfi, code-ll, code-awkward-estates, code-willscan, code-claude-code-design-system, openinherit/code-standard (with strong stealth-discipline banner), openinherit/code-backoffice
  • Added staleness + deltas banner to code-mfi/docs/superpowers/specs/mfi-design-spec.md v2.3
  • Updated docs-strategy/README.md v1.0 → v2.0 with canonical doc pointers + sister-repo map

Phase F — Archival + gold-preservation retrofit

  • Moved tt-products/mfi/v4/ + v4.1/ + upgrade-prompt.md (9,243 lines) to docs-strategy/_archive_/legListMFI-pre-split/ per Rich’s #11
  • Moved tt-products/henry/henry-gap-analysis.md to docs-strategy/_archive_/henry-shelved/ per Rich’s #10
  • Added archive README + per-folder ARCHIVED.md banners
  • Retrofitted 4 canonical docs with new gold-preservation fields: brand-architecture.md v1.5 + option-G.md v1.3 + scorecard v1.8 + roadmap v1.6

Canonical docs now in place (read these before any TT work)

  1. ~/testatetech/docs-strategy/brand-architecture.md v1.5 — CANONICAL brand architecture
  2. ~/testatetech/docs-strategy/frontmatter-conventions.md v1.0 — filing-system discipline
  3. ~/testatetech/docs-strategy/CLAUDE.md — strategy-repo session context
  4. ~/testatetech/docs-strategy/docs/superpowers/specs/2026-04-19-inherit-v2-roadmap.md v1.6 — 24-month roadmap
  5. ~/testatetech/docs-strategy/docs/superpowers/specs/2026-04-18-inherit-v2-architecture-options-scorecard.md v1.8 — architecture scorecard (best-current-proposal, not approved)
  6. ~/testatetech/docs-strategy/docs/superpowers/specs/2026-04-18-inherit-v2-architecture-options/option-G.md v1.3 — current-best INHERIT v2 proposal
  7. ~/testatetech/docs-strategy/ideas/ — tangent-gold capture (5 files + README + template)

Key commits (in order)

  • 331da21 — canonical brand-architecture v1.0 + IW-first reorder + Option G+ status revert
  • 61cda1b — brand-architecture v1.1 + roadmap v1.2 — InheritKit built before IW started
  • c1f2ac0 — brand-architecture v1.2 + roadmap v1.3 — OpenInherit (next version) first in build order
  • 38bbea8 — brand-architecture v1.3 + roadmap v1.4 — LL v1 removed from chain + “IW-first” dropped from title
  • 2d006c3 — brand-architecture v1.4 + roadmap v1.5 — OpenInherit public-release plan (gradual + launch-coupled)
  • 7d8fd01 — frontmatter-conventions v1.0 + ideas/ folder + per-brand CLAUDE.md embedding
  • b0e63bb 43c6d15 4cedabe 873948b 9d7fee9 47e475f — per-brand CLAUDE.md commits across code-mfi, code-ll (rebased onto Josh), code-inheritwills, code-awkward-estates, code-willscan, code-claude-code-design-system
  • 19ae973 86f571d — openinherit CLAUDE.md commits (stealth discipline)
  • plus one coming: archive + gold-preservation retrofit

Memories saved tonight

  • project_openinherit_release_discipline.md — stealth-discipline rules (don’t improve preemptively; launch-coupled events; partner NDA access)
  • feedback_ai_raises_delegation_bar.md + project_nova_partnership_deprioritised.md — earlier but reinforced
  • project_inheritwills_mfi_subscription.md — IW subscription-native + MFI Level 0 Monitoring
  • project_session_2026_04_19_evening.md — this file

What’s genuinely decided vs what’s still open

DECIDED tonight (absorbed into canonical docs):

  • InheritWills is B2B + D2C (Notion-pattern, one brand one codebase)
  • InheritWills launches first as consumer brand
  • InheritKit.com launches next (commercial + partner portal + Advisory BackOffice)
  • Build order: OpenInherit → InheritKit → IW → MFI/LLv2 parallel
  • LL v1 (Josh) is independent track — don’t interrupt
  • LL v2 (parallel copy) is a separate experimental project
  • Henry shelved (ideas → InheritKit)
  • WillScan.ai not launching standalone
  • LegListMFI formally archived
  • CherishedItems very low priority
  • WillsInFaith follows MFI by ~2 months per jurisdiction
  • AwkwardEstates is content + lead-gen
  • Nova partnership deprioritised
  • Advisory BackOffice lives on InheritKit.com (not MFI)
  • Will scanner only on InheritWills
  • OpenInherit public-release: gradual + launch-coupled + partner-first + stealth discipline
  • Subscription model: IW subscription-native + MFI opt-in Level 0 Monitoring (preserves §34 one-off trust for Levels 1-3)
  • Option G+ v1.2 is current best INHERIT v2 proposal but NOT approved (pending product requirements)

STILL OPEN (surfaced in canonical docs’ open-questions):

  • Catalogue position (6th family or sub-family of Assets?)
  • INHERIT-Core separate module or shared kernel?
  • Per-family partner model shape
  • LLv2 parallel-copy project scope + repo
  • Jurisdictional strategy per brand
  • Governance evolution beyond TT
  • Will Power brand-name decision (deferred; ideas/ file preserved)
  • IW subscription pricing validation
  • MFI spec v2.3 realignment timing

How to apply in future sessions

  • Read brand-architecture.md v1.5 + frontmatter-conventions.md v1.0 FIRST before any TT work
  • Do NOT improve openinherit.org preemptively (stealth discipline)
  • Do NOT realign MFI spec v2.3 without Rich’s explicit go-ahead
  • Do NOT propose Nova as build partner
  • Don’t interrupt Josh’s LL v1 build
  • When capturing a new tangent idea: ideas/YYYY-MM-DD-slug.md with full frontmatter
  • When superseding a doc: explicit supersedes + superseded_by chain
  • When touching canonical docs: retain gold-preservation fields (origin, related, open-questions, gold-value)
  • When in doubt about status value: consult frontmatter-conventions.md §4 (11-value canonical vocabulary); don’t invent new values silently

Graph View

Backlinks