Go / No-Go Memo

Question: Would you allocate $5M tomorrow?
Answer: NO-GO (unchanged for $5M).

One-liner: Engineering controls improved materially in May 2026; P0 hardening added live-trading gate, decay quarantine v1, run manifest, P0-7 standalone qgtm-watchdog.service, and Prometheus decay/AISC gauges — but economic validation remains incomplete until GAP-002 covers all 52 sleeves in backtest_results/ (batch in progress).

Rationale

Dimension	Assessment
Edge	Quarantine cohort has negative OOS; remaining sleeves lack checked-in PBO/DSR
Risk	Kill-switch, compliance, drawdown code is credible on `main` — verify deploy SHA
Execution	TCA wired; still ETF-centric vs stated GC/SI mandate
Learning	Partial — decay auto-quarantine v1 in daemon; full ML loop still isolated in `dead_ml_loop/`
Operations	Self-hosted deploy + CI gate; standalone DMS watchdog on droplet; operator UI improved (#74–#77)

Institutional readiness (honest): ~63% — up from ~58% after P0-7 deploy path + observability; still blocked on full GAP-002 coverage and $5M economic proof.

Conditional GO paths

Capital	Requirements
$500k paper continue	GO — current posture
$1M live seed	GAP-002 done + 90d paper + live drills + `qgtm_live_trading_enabled` review
$5M	Above + TCA < 15 bps + realization > 0.7 + futures or explicit ETF capacity sign-off
$25M	Dedicated execution + second broker + 12m track
$100M	Program-level PM, compliance, ADV model, GC/SI block market

Roadmap summary

flowchart LR
  A[Today: Paper 46 sleeves] --> B[GAP-002 + kill list]
  B --> C[60d paper proof]
  C --> D[1M live seed]
  D --> E[5M scale]
  E --> F[25M futures + TCA]
  F --> G[100M program]

Sign-off (unfilled)

Role	GO/NO-GO	Date
PM / Allocator	NO-GO	2026-05-17
Risk
Eng
Operators (user, Naz)