sdc-strategy-filter-focus-brainstorm

SDC Filter+Focus — Real Brainstorming Pass

Background

Spun out from sdc-dept-scaffold-mbr-absorb (2026-08-28). That task ported SDC/Strategy/Filter+Focus.md’s structure from MBR and seeded it with a first-pass draft — the known rocks already named in Core/DASHBOARD.md and the transferables inventory, not a real brainstormed list. MBR’s own Filter+Focus.md (73+ ranked rocks) took a dedicated brainstorming session (AI Studio + Claude Code, April 2026) to build; SDC hasn’t had that session yet.

Tasks

- Talbot extracts SDC ideas from Dynalist into KB-OS. Many SDC ideas currently live only in Dynalist, not yet captured in the vault for ranking. This is a manual extraction step — Claude has no Dynalist access.
- Apply the Filter+Focus process to SDC’s extracted ideas. Rank against SDC/Strategy/Filter+Focus.md’s existing first-pass draft; update both SDC/Strategy/Filter+Focus.md and, where a rock is big enough, Core/DASHBOARD.md’s vault-wide Biggest Rocks table.

Success Criteria

SDC/Strategy/Filter+Focus.md reflects a real brainstormed rock list (MBR’s shape: All Rocks Ranked / By Category / By Phase), not just the first-pass seed.
Any rock big enough to matter cross-business is reflected in Core/DASHBOARD.md.

Notes

No rush — “later, when appropriate priority/rock” (Talbot, 2026-08-28). focus: 3_later.
/filter-focus-refresh skill exists but is currently hardcoded to MBR’s Filter+Focus.md — extending it to SDC is a natural follow-on once this file has real content, not a prerequisite for this task.

Context (Auto-Prepared) — 2026-08-28

Resolved File Paths

SDC/Strategy/Filter+Focus.md → /mnt/d/FSS/KB/SDC/Strategy/Filter+Focus.md (verified: exists — the first-pass draft this task upgrades)
MBR/Strategy/Filter+Focus.md → /mnt/d/FSS/KB/MBR/Strategy/Filter+Focus.md (verified: exists — the structural model, 73+ ranked rocks, 3 formats: All Rocks Ranked / By Category / By Phase, plus a Key Decision Log)
Core/DASHBOARD.md → /mnt/d/FSS/KB/Core/DASHBOARD.md (verified: exists — vault-wide Biggest Rocks roll-up table to update if a resulting rock is cross-business-significant)
Dynalist itself is external, not in the vault — no path to verify; this is why subtask 1 is Talbot-only.

KB Context

SDC’s current Filter+Focus.md states plainly it’s a “structural port, not an exhaustive brainstorm” — seeded from Core/DASHBOARD.md’s existing Biggest Rocks table and the MBR→SDC transferables inventory, not a real ideation session.
MBR’s version took a dedicated multi-session brainstorm (AI Studio + Claude Code, refreshed April 2026) to reach its current depth — that’s the bar this task is closing SDC up to, not a mechanical port.
Ranking criteria to reuse (already stated in SDC’s draft, ported from MBR): Scalability (low human involvement at scale) × Growth Impact × Ease of early execution. Framework: James Clear A→B→Z.

Dependencies

Prerequisite status: all clear (none listed)
External systems needed: Dynalist (Talbot’s own access — not available to Claude)
Files that must exist before execution: SDC/Strategy/Filter+Focus.md (exists, is the file being upgraded)

Execution Notes

Model recommendation: subtask 1 (Dynalist extraction) is Talbot-only, no model needed. Subtask 2 (applying the Filter+Focus ranking process to the extracted ideas) is a judgement-heavy strategy task once ideas exist — Opus, matching the model used for the parent dept-scaffold task’s own strategy work and MBR’s original brainstorm sessions.
Assignee: talbot (subtask 1 is manual; subtask 2 can be handed to an agent once subtask 1 lands)
Key constraints: don’t fabricate SDC rock content that isn’t grounded in Talbot’s real ideas or existing vault content — NEVER GUESS applies directly here, since a fabricated “big rock” would misdirect real priority.