MIS CROSS-SHEET LOGIC ANALYSIS — Does the math actually fit?

🧮 THE SACS DISCREPANCY EXPLAINED — math is consistent

Conclusion: v1 and Sandbox are NOT showing different conviction levels — they're showing the same underlying conviction at two different points in an upgrade chain. Sandbox is the testbed before Fix #5; v1 is after. Both are valid; v1 is the next step.

Production's scale is genuinely different — uses Sam's older 13-weight Composite (0-55 effective range) with tristate Flow (28/44/76 only). It's a different lineage.

📐 ALL 8 FIXES IN runSACSUpgrade.js — what they do, what they affect

Action: Run runSACSUpgrade from v1's MIS Tools menu. The function logs to SACS_Upgrade_Log tab showing which fixes are DONE / SKIPPED / ALREADY_APPLIED. That single run answers all the "Unknown" cells above.

🔍 LOGIC INCONSISTENCIES SURFACED

Inconsistency #1 — v1 says "ACTIONABLE TODAY: 0" but should say "5 candidates"

v1's email I patched Friday surfaces ACTIONABLE = Lean In + In Sync. Currently shows 0. But v1's Top Conviction shows ABBV/LLY/AMT/GD/PEP all as Lean In setups with SACS 66-86. The bottleneck is the actionFlag === '✅ In Sync' filter. The actionFlag is defined as ≥42 In Sync per legend. After Fix #5's 1.8× expansion, scores like 86.77 should EASILY clear actionFlag — but the actionFlag uses the Composite (not the post-expansion SACS) so it's still the old scale.

Verdict: Fix #5 expanded SACS but didn't expand the actionFlag threshold. Asymmetric upgrade. Need to either (a) lower actionFlag threshold, or (b) compute actionFlag from post-expansion SACS, or (c) don't tie Actionable to actionFlag at all — tie it to FinalState=ADD from Snapshot's 5-gate pipeline directly.

Inconsistency #2 — Test_Entry tab present but blank in V11 Microservices

You asked specifically about Test_Entry. It exists in Architect V11 Microservices (1MU_Ni) as Test_Entry tab. Content per Drive snippet:

📱 MIS MOBILE SANDBOX
ENTER TICKER:
Category, No usable data / invalid ticker
SACS Score
Live Price
Model Size
Stop Loss
Target
Downside %
Upside %
Risk $
Target $
Reason
Blocks
Headline Check

It's a SCHEMA only — placeholder cells with no live data. Never wired. Per the Mobile Command Center vision Sam discussed with Gemini, this was supposed to be the "type ticker → get answer" mobile pattern. It exists; it's not connected.

Inconsistency #3 — Mobile Command Center: code exists, no sheet uses it

Sam dropped scripts/mobile_command_center.js today — checkbox-driven controls for Dashboard tab cells Z1-Z5. Zero sheets currently have these checkboxes. The code is a DESIGN, not a deployment. Adding it requires:

1. Paste mobile_command_center.js into v1's Apps Script
2. Set up onEdit trigger: handleMobileButtons
3. Manually add 5 Insert→Checkbox cells in v1's Dashboard tab at Z1, Z2, Z3, Z4, Z5
4. Label cells in column AA next to each checkbox

~10 minutes of setup. Not done yet.

Inconsistency #4 — Architect V11 Microservices SACS uses different model

The Microservices Playbook_Dashboard shows DUOL/IBIT/GD with SACS 100/100/97. SACS=100 doesn't appear anywhere else. v1 max was 86.77 (LLY). Production max was 50.99 (INTC). Either:

• Microservices uses raw component sum without normalization (likely — produces 100s easily)
• OR uses post-expansion + further capping
• OR uses a different formula entirely

Without reading its bound script 1mSbzO directly (which the runSACSUpgrade context implies is older code), can't confirm. This is the third independent SACS scale. Three in production = bad.

Inconsistency #5 — Trade_Log formatting bugs across multiple sheets

v1's Recent Trade History shows: 46125, 93947.00%, 95519.00% — that's date as serial number, prices as percentages. Same bug Sam's V11 teardown flagged. Inherited from production. Cell format issue, not formula. 2-minute fix in each sheet's Trade_Log tab.

📸 IOS SCREENSHOTS — what Sam was looking at on phone Apr 28

Reviewed one of the 14 screenshots in the new screenshots/ subfolder. They show the MIS Morning Brief email rendered on phone — Mag 7 panel with prices, percentages, prev close + 30-Day Charts (mentioned tickers) with embedded TradingView/Yahoo chart links. This is Sam's reference visual for what the email should look like. The current v1 email already has both sections. Visual format is consistent.

📂 MIS FOLDER ORGANIZED — what's where now

🚨 RECOMMENDED HARD DECISIONS (do these in order)

Decision 1 — Kill the duplicate-emails source NOW

Open script 1mSbzO (Architect V11 Microservices) → Triggers tab → DELETE ALL TRIGGERS. That stops the weekend "MIS Power Hour" emails. 60 seconds.

Decision 2 — Run runSACSUpgrade on v1 NOW

v1 script 1yBMzt → paste runSACSUpgrade.js into a new Apps Script file → save → from MIS Tools menu run "Run SACS Upgrade (V2)" → check SACS_Upgrade_Log tab to see what changed. 5 min. This applies all 8 fixes including the critical RiskGate fix that unlocks LLY/AMD/MRVL.

Decision 3 — Move v2/Sandbox to graveyard

Per workspace policy (CLAUDE.md): rename in Drive, don't trash. Suggestions:

• v2 (1F31z) → MIS_DEAD_excel-roundtrip_2026-05-03
• V2 SACS Sandbox (1HtMOd) → keep IF you want a comparison testbed for the pre-Fix-5 SACS, otherwise rename to MIS_SANDBOX_pre-fix5_2026-05-03

Decision 4 — Pick primary engine

Recommendation: Option B with strict sequencing — (1) rotate credentials, (2) port v1's runSACSUpgrade + email patches into production, (3) retire v1 to sandbox. This preserves your 200+ logged runs of intelligence and the Schwab pipeline while gaining the v1 improvements.

Decision 5 — Test_Entry / Mobile Command Center

If you want phone-driven control: paste mobile_command_center.js into chosen primary script + add 5 checkboxes in Dashboard. ~10 min. The Test_Entry tab in Microservices is a placeholder — wire it up only after primary engine is chosen.

📊 LOGIC FIT VERDICT — does it all hang together?