Skip to content
FSS Knowledge Base

$MART DEBT Strategy Library v0.1 — the risk-first knowledge layer

Section titled “$MART DEBT Strategy Library v0.1 — the risk-first knowledge layer”

Background

Section titled “Background”

Strategic-Plan v2.0 Phase 1b. The ChatGPT brainstorming session converged on a single structural idea worth taking: the fundamental unit of the business is not an article, video, or chapter — it is a strategy, as a versioned record with a standard schema, sorted first by risk. Talbot had already independently decided to build a library of all debt strategies sorted on the primary axis of risk (session turn 3).

That library is the foundation everything else in the plan sits on: it is what the F.A.S.T.T. micro-apps apply, what a future Navigator navigates, what an advisor cites, and what an AI search engine finds and quotes. It is also the one Phase 1 workstream that needs no credentials, no deploy, and none of Talbot’s writing time.

Why v0.1 is deliberately small: the session proposed community pull requests, contributor reputation, strategy maintainers, red teams, and bug bounties. Per Brainstorming Upgrades #6, none of that is in scope — a solopreneur has no community to moderate and no maintainer bench. v0.1 is a published, non-interactive reference. Nothing more.

Sources that actually exist (do not invent beyond these — see Success Criteria):

  • ~/projects/monorepo/packages/sd-math/ — three fully ported and verified strategies: interest-only loan, term loan, RRSP catch-up loan
  • ~/projects/monorepo/packages/sd-math/docs/strategy-notes/smith-manoeuvre.md and debt-swap.md — deferred, notes-only
  • ~/projects/monorepo/packages/sd-math/docs/design.md, M5-M6-plan.md, accuracy-audit.md
  • Better-Rates-Strategy — the zero-risk first rung
  • Smart Debt Offerings — course-content section (interest-only→term-loan combinations, Buy More Low)
  • SDC/Strategy/IP/ and Identity/Internal View/Experience.md — the publication lineage

Tasks

Section titled “Tasks”
  • Define the strategy record schema and write it once as SDC/Strategy/Library/SCHEMA.md. Fields, from the session’s turn-3 proposal, filtered to what can actually be filled today: name · objective · risk classification · prerequisites · mechanism · benefits · risks · failure modes · costs · tax considerations · who it may suit · who should avoid it · implementation outline · evidence status · counterarguments · variants · related strategies · sources · version · last reviewed · open questions.
  • Define the risk axis explicitly — the primary sort key, and the one that governs Rule 1 of the Strategic Plan. It must support the CEO’s own three-line standard: everyone should ACT on negative-risk strategies; most should consider little-risk ones; most should NOT act on modest- or high-risk ones. Proposed tiers, to be justified not assumed: negative-risk (strictly better off, e.g. refinancing to a lower rate) · zero-risk · low-risk optimization · moderate · risk-magnifying (any form of borrowing to invest). Cross-check against SDC/Risks/JOB_DESCRIPTION.md and reconcile with sdc-behavioural-solutions-debt if that task has landed.
  • Add the secondary axes the session proposed, only where they can be filled from evidence: complexity · capital required · liquidity impact · reversibility · time to benefit · evidence strength · behavioural difficulty. Drop any axis that cannot be populated for most records — an axis with 3 of 25 values filled is noise.
  • Write 20–30 strategy records into SDC/Strategy/Library/, one file per strategy. Start with everything already verified in sd-math (highest confidence), then the notes-only strategies, then the rest.
  • Write SDC/Strategy/Library/Library.md — the index: every strategy in one table, sorted by risk tier, with objective and evidence status.
  • Add F.A.S.T.T. to Core/Misc/Glossary.md — Focused, Adaptive, Simple, Tailored, Trustworthy. The glossary currently records F.A.S.T. (four terms); the fifth was added 2026-09-06 and is now the design standard the library’s evidence status field serves. Flag, do not edit, the design-context skill in ~/ai-config/ — that deploys globally and is Talbot’s call.

Success Criteria

Section titled “Success Criteria”
  • Every claim in every record traces to a named source — a sd-math implementation, a repo strategy note, a KB file, or a cited external source. A record with no source for a claim marks that field unverified — needs Talbot, and does not guess. This is the single most important criterion: a fabricated financial strategy record is worse than a missing one, and this library is intended to be publicly citable.
  • Every record carries a risk classification, and the classification is justified in one line, not asserted.
  • ≥20 records published; the index table renders correctly in Obsidian.
  • The schema is written once and every record conforms to it — no per-record improvisation.
  • No community, contribution, or reputation mechanism is built (explicitly out of scope).

Notes

Section titled “Notes”
  • F.A.S.T.T. — Focused, Adaptive, Simple, Tailored, Trustworthy. The product-design standard.
  • RRSP — Registered Retirement Savings Plan (Canada). A catch-up loan borrows to fill unused contribution room.
  • Smith Manoeuvre — a Canadian strategy converting non-deductible mortgage debt into deductible investment debt as the mortgage is paid down.
  • SSOT — Single Source of Truth.
  • Jurisdiction is a required field, not a footnote. The ported strategies are Canadian-tax-modelled, and since the 2026-09-06 Canada-first decision, Canada is the first market — so Canadian records are the priority, not the exception. Never silently present a Canadian strategy as applicable to a U.S. client; mark U.S. applicability unverified until sdc-risks-us-compliance-research runs (deferred until the Canada restart is complete).
Section titled “Related”

Claude Response — 2026-09-07

Section titled “Claude Response — 2026-09-07”

Summary:

  • 25 strategy records published in SDC/Strategy/Library/, one file per strategy, all conforming to one schema. Above the 20-record floor, and every record is backed by a named source rather than padded to reach the count.
  • SCHEMA written — the record template, the risk axis, the secondary axes, and the mapping from tier to Rule 1’s permitted intervention.
  • The risk axis was redesigned, not just written down. Four different tier vocabularies exist across SDC documents (this task file’s five, Strategic-Plan Phase 1b’s five, Rule 1’s three, SDC/Risks/JOB_DESCRIPTION.md’s three) and they do not agree. The structural defect: “risk-magnifying” is a mechanism label while “moderate/high” is a magnitude label — not points on one axis, so the scale could not be sorted. Split into one sortable magnitude scale R0–R4 plus a separate leverage flag, with an explicit R0–R4 → Rule 1 mapping table (SCHEMA §2.4). Nothing lost; the primary sort key now works.
  • Library index written — every record in one table, sorted by tier, with objective, leverage flag, jurisdiction and evidence status. Table integrity verified programmatically (0 column mismatches) and all 25 wikilinks resolve.
  • Secondary axes: 5 kept, 2 dropped, with reasons. Kept complexity · capital required · liquidity impact · reversibility · time to benefit. Dropped evidence strength (duplicates the Evidence status field) and behavioural difficulty (cannot be populated from evidence until sdc-behavioural-solutions-debt runs — filling it now would be guessing on exactly the dimension that task exists to settle).
  • The sourcing rule held. 3 records implemented-and-verified (the sd-math ports), 15 external-sourced or documented, 3 design-note, 2 unverified stubs. unverified — needs Talbot was used as a field-level marker, never as licence to invent a record.
  • The index states what the library cannot do. Only 3 of 25 strategies can currently be computed — the rest publish a mechanism with no outcome. That is in the index’s own table, not buried.
  • Pointers added so nothing duplicates: Better-Rates-Strategy keeps the positioning and now points at its four mechanism records; SDC/Strategy/ROADMAP.md lists the Library under What Exists.
  • Committed to the vault repo.

The glossary task was already done — the premise was stale. Core/Misc/Glossary.md has carried a full F.A.S.T.T. entry since 2026-09-06 (Focused · Adaptable · Simple · Tailored · Trustworthy), alongside the mBR-scoped F.A.S.T. Two notes: the fifth term is Adaptable, not “Adaptive” as this task file says — Strategic-Plan agrees with the glossary, so the task file is the outlier. And the design-context skill in ~/ai-config/ is F.A.S.T.-only throughout (SKILL.md, references/principles.md, voice.md, target-decks.md, target-reports.md) — flagged, not edited, per the instruction; it deploys globally and is your call.

Two records are honest stubs, published rather than omitted:

  • buy-more-low — named core IP, referenced across four SDC documents, with the Buffett quote and two named data points. No mechanism exists anywhere in the vault; SDC’s own note says “Review all ideas in ‘Buy More Low’ Wealth Accelerator notes” and those notes are not here.
  • cash-damming — named by Robinson Smith as a Smith Manoeuvre accelerator; mechanics not sourced.

Both say so plainly in every field. The alternative was to reconstruct them from general knowledge, which is the one thing the Success Criteria forbid.

Next Steps for Talbot:

  • Sign off the risk axis (R0–R4 + leverage flag), or correct it. This is the load-bearing decision — every record inherits it, and the product’s Rule 1 behaviour keys off it. The Strategy JD is A1 (you approve) and routes any change to how leverage strategies are presented through Risks / Risk-Challenger first. SCHEMA §2 is written as a draft awaiting exactly that. *
  • R1 has exactly one record — keep five tiers or collapse to four? eliminate-high-cost-consumer-debt-first is the only R1, and R0 and R1 map to the same Rule 1 behaviour, so nothing in the product turns on the split. They are kept apart only because your own standard and the Risks JD both name negative-risk and zero-risk as separate things. My read: keep, with the distinction stated as descriptive (SCHEMA §2.2 now says so). Collapsing to a single negative/zero tier is the clean alternative. *
  • Where are the “Buy More Low” Wealth Accelerator notes? Everything about BML is blocked on them, and it is named core IP inside $MART DEBT Wealth. *
  • Keep or drop the two unverified stubs (buy-more-low, cash-damming). Publishing them shows the gap; omitting them hides it. My read: keep, because a public library that visibly says “we don’t know yet” is the F.A.S.T.T. Trustworthy claim working. *
  • Three scope calls, each flagged inside its record. (a) Embedded-leverage products — do leveraged-etfs and return stacking belong in a debt library at all? (b) Income splitting — prescribed-rate-spousal-loan is a loan whose terms are the strategy, but it is tax planning. (c) Corporate — corporate-borrowing-to-invest for an advisor audience serving incorporated clients. *
  • Render-check Library in Obsidian. Table column integrity and all 25 wikilinks were verified programmatically, but no Obsidian CLI is installed on this machine, so actual render is the one Success Criterion I could not close myself. *
  • design-context skill: F.A.S.T. → F.A.S.T.T. for SDC targets. Flagged only — global deploy, your call. *

Path to Completion

Section titled “Path to Completion”

Review it: start at the index, then one record of each confidence class — /mnt/d/FSS/KB/SDC/Strategy/Library/Library.md · D:\FSS\KB\SDC\Strategy\Library\Library.md. Highest confidence: interest-only-investment-loan.md. Weakest, deliberately: buy-more-low.md. The axis decision itself: SCHEMA.md §2. No /create-review-guide — that produces a repo docs/Review-*.md for a dev project; this deliverable is vault markdown you read in Obsidian.

Project SSOT: /mnt/d/FSS/KB/SDC/Strategy/Strategic-Plan.md · D:\FSS\KB\SDC\Strategy\Strategic-Plan.md — Phase 1b is this workstream · /mnt/d/FSS/KB/SDC/Strategy/ROADMAP.md — thin index, now lists the Library. Oversight SSOT: Strategic-Plan.md.

Now: /task-continue D:\FSS\KB\SDC\Strategy\Tasks\sdc-strategy-library-v01.md Run it after you answer the risk-axis checkbox above. It handles either answer: axis changed → round 2 rewrites SCHEMA plus every affected record in one pass (the largest edit in this task, not a close-out); axis approved as drafted → round 2 confirms no change and hands straight to /task-complete.

Then, in order:

  1. /task-complete D:\FSS\KB\SDC\Strategy\Tasks\sdc-strategy-library-v01.md — closes v0.1 once the axis is signed off and the stub decision is made.
  2. /task-start D:\FSS\KB\SDC\Strategy\Tasks\sdc-behavioural-solutions-debt.md — the direct successor. It settles the intervention ceilings SCHEMA §2.4 currently defaults to the conservative reading of, and unblocks the behavioural difficulty axis this round had to drop.
  3. /task-start D:\FSS\KB\SDC\IT\Tasks\sdc-sdapp-trust-receipt.md — Phase 1a. Pairs with the library: the receipt is what makes evidence-status visible in the product rather than only in the vault.

Related open tasks (parallel work, not choices):

  • sdc-behavioural-solutions-debt — startable now (prerequisite sdc-dept-scaffold-mbr-absorb is complete) — /task-start D:\FSS\KB\SDC\Strategy\Tasks\sdc-behavioural-solutions-debt.md
  • sdc-sdapp-trust-receipt — startable now, no prerequisites — /task-start D:\FSS\KB\SDC\IT\Tasks\sdc-sdapp-trust-receipt.md
  • sdc-risks-canada-leverage-compliance — active, round 1 done — /task-continue D:\FSS\KB\SDC\Risks\Tasks\sdc-risks-canada-leverage-compliance.md. Touches this task: its §4 boundary decides whether a per-client library record crosses into regulated advice.
  • sdc-mktg-canada-restart — active, round 1 done — /task-continue D:\FSS\KB\SDC\Mktg\Tasks\sdc-mktg-canada-restart.md. Its message drafts can cite library records once the axis is signed off.

Blocked on you (no command fixes these):

  • Risk-axis A1 sign-off, with Risks / Risk-Challenger review first — per the Strategy JD.
  • The “Buy More Low” Wealth Accelerator notes — nothing in the vault reaches them.
  • Obsidian render-check of Library.md — no Obsidian CLI on this machine.
  • design-context skill F.A.S.T. → F.A.S.T.T. — global deploy.

Done when: the risk axis is CEO-approved (and Risks-reviewed), the two stub records are kept or dropped by decision, and Library.md renders correctly in Obsidian — at which point v0.1 is a publishable reference and Strategic-Plan Phase 1b’s library condition (“≥20 strategies published with risk classifications”) is met.

Talbot Response

Section titled “Talbot Response”

Claude Response — 2026-09-21 (hand-off, not a round)

Section titled “Claude Response — 2026-09-21 (hand-off, not a round)”

The artifact this task built has left the department

Section titled “The artifact this task built has left the department”

On 2026-09-20 the CEO decided that the strategy library is intellectual property, not strategic planning, and created SDC/IP/ to hold it. Acted on 2026-09-21 by SMART-DEBT-Strategies-Library (now in SDC/IP/Tasks/):

  • SDC/Strategy/Library/ → SDC/IP/Strategies/. Same files, same wikilinks — short-form links resolve by filename, so nothing broke; a new department owns them.
  • The risk axis this task’s Next Steps were blocked on is decided. tier: R0–R4 is replaced by a signed risk-level: -1 … 4 — negative risk at -1, zero risk (“Can’t Lose”) at 0, then low / moderate / higher / highest. All 25 records migrated; the §2.4 Rule 1 mapping was carried across row by row, not re-derived.
  • Schema v0.2 shipped — 10 fields added, id and confidence rejected, code added as a Chart-of-Accounts-style alias. SCHEMA is still the SSOT and now lives under SDC/IP/.
  • The Risks (Risk-Challenger) gate moved with the library and is restated verbatim in SDC/IP/JOB_DESCRIPTION.md, so the move did not silently drop it. The Risks review of the axis is still outstanding — it did not happen here, and the CEO’s decision on the scale does not close it.
  • strategy-lint now enforces schema conformance mechanically — ~/utils/strategy-lint/strategy-lint lint, registered at Core/IT/Utils/Custom/strategy-lint.md.

Do not edit SCHEMA.md or any record from this task. That work transferred to SDC/IP/. This task’s own round 2 was staged to “rewrite SCHEMA plus every affected record in one pass” — that pass has run, from the IP side.

What is still genuinely this task’s to close

Section titled “What is still genuinely this task’s to close”

Three items are Strategy-dept decisions the move did not resolve:

  • The two unverified stub records (buy-more-low, cash-damming) — keep as honest stubs, or drop. *
  • The “Buy More Low” Wealth Accelerator notes — nothing in the vault reaches them, so buy-more-low stays a stub until they surface. *
  • The three scope calls — embedded-leverage products, income-splitting, corporate. *

Path to Completion

Section titled “Path to Completion”

Review it: the migrated library at its new home — /mnt/d/FSS/KB/SDC/IP/Strategies/Library.md · D:\FSS\KB\SDC\IP\Strategies\Library.md. Until the move command runs (it is the first Next Step on the IP task), these files are still at SDC/Strategy/Library/.

Project SSOT: /mnt/d/FSS/KB/SDC/Strategy/Strategic-Plan.md · D:\FSS\KB\SDC\Strategy\Strategic-Plan.md — Phase 1b. Oversight SSOT: Strategic-Plan.md.

Now: /task-continue D:\FSS\KB\SDC\IP\Projects\Strategies-Library\Tasks\sdc-ip-mdw-strategy-foldin.md The library work continues there, not here. (Repointed 2026-09-23: the IP task SMART-DEBT-Strategies-Library became the ongoing project Strategies-Library.)

Then, in order:

  1. /task-complete D:\FSS\KB\SDC\Strategy\Tasks\sdc-strategy-library-v01.md — close this once the three items above are answered. Phase 1b’s condition (≥20 strategies published with risk classifications) is already met.

Blocked on you (no command fixes these): the two stub records, the BML notes, the three scope calls, and the outstanding Risks review of the risk axis.

Done when: the three open items are answered and this task closes — the library itself is no longer this department’s work.

Talbot Response

Section titled “Talbot Response”