# System-level review

Date: 2026-05-16

Per-property assessment + recommendation. 13 active properties + 1 inactive.

---

## foundation — **CORE_INFRASTRUCTURE**

| metric | value |
|---|---|
| technical debt score | HIGH (the 6 Phase 0 deliverables literally exist to repay this) |
| business leverage | HIGHEST (compounding-infra lives here) |
| operational confusion | MEDIUM (some risk_hud/identity-fn dup) |
| duplication risk | LOW (centralizing reduces it) |

**Recommendation: CORE_INFRASTRUCTURE.**

6 of 7 compounding-infra tasks are in or directly support foundation. The Phase 0 freeze is doing its job — until p0-freshness-dashboard, p0-orchestrator, p0-identity-fn, p0-daily-summary-view, p0-param-surface-registry land, every downstream property is paying recurring cost. Saturday autonomous work should target foundation first.

---

## scorecards — **STABILIZE**

| metric | value |
|---|---|
| technical debt | HIGH (27 open; 5 cross-property tasks living here per drift) |
| business leverage | HIGH (Scorecards is where James grades trades) |
| operational confusion | HIGH (highest pending count, lowest velocity among big props) |
| duplication risk | HIGH (catalog drift documented; 3 strategy-* tasks belong elsewhere) |

**Recommendation: STABILIZE.**

Don't pile on new scorecard features. Three actions stabilize it: (1) re-home the 5 mis-tagged tasks per drift.md, (2) close the webhook enrichment chain to retire all the "N/A placeholder" debt, (3) ship the catalog refactor so classify_entry() and _scientific_overlay.js stop diverging. After those, scorecards is positioned to absorb phase-4 multi-trader expansion cleanly.

---

## accountability_hud — **STABILIZE** (under HUD-off-limits)

| metric | value |
|---|---|
| technical debt | LOW (v2 is recent + well-documented in BUILT.md) |
| business leverage | HIGHEST (live cockpit James uses daily) |
| operational confusion | LOW (clear ownership: jimbo) |
| duplication risk | MEDIUM (risk_hud cluster overlaps) |

**Recommendation: STABILIZE.**

Two open P0 bugs (risk-bump + ticket-meter) on the DEV HUD need jimbo's resolution before any new feature work. The 4 HUD card requests (Trader State / Composure Trend / Mood / Live Grade) are HUD-off-limits per memory; supervised only. Don't autonomously touch the live cockpit. The dedup decision (risk_hud vs acct-v2) should land before risk_hud build resumes.

---

## mgp — **STATUS_QUO** (with watch on scheduler)

| metric | value |
|---|---|
| technical debt | LOW (independent pipeline cutover imminent) |
| business leverage | HIGH (every trading day starts here) |
| operational confusion | MEDIUM (cron drift documented 2 days running) |
| duplication risk | LOW |

**Recommendation: STATUS_QUO** with one P0 watch item.

The independent MGP pipeline (`mgp-drop-vital-independent-source`, in_progress) is the highest-attention work happening on the project; Saturday is the right day to cut it over per today_focus. But the scheduler drift documented in `mgp-cron-skipped-5-14` is a separate P0 that gates the cutover safety — if the scheduler fires the wrong slot Monday morning, the cutover looks broken. Resolve the scheduler drift first.

Other MGP tasks (multi-tf bias, breadth indicator click-through, rsi-overextended refinement) are quality improvements that can land in any order post-cutover.

---

## param_surface — **STABILIZE → AUTOMATE**

| metric | value |
|---|---|
| technical debt | MEDIUM (VWAP grid only; no registry yet) |
| business leverage | HIGH (the only quantitative edge surface) |
| operational confusion | LOW |
| duplication risk | MEDIUM (will compound if multiple strategy grids built without registry) |

**Recommendation: STABILIZE in the short term; AUTOMATE after registry lands.**

`p0-param-surface-registry` is the gate. Once that ships, every new strategy gets its own grid by adding a JSON row. Until then, building a second strategy grid means duplicating the VWAP pipeline — high duplication risk.

---

## strategy_library — **STABILIZE**

| metric | value |
|---|---|
| technical debt | LOW (page is static; canonical catalog migrated 5/15) |
| business leverage | HIGH (the entry-point doctrine surface) |
| operational confusion | MEDIUM (4 P0/P1 + 11 open, lowest velocity 0.27) |
| duplication risk | HIGH (catalog inlined in 5 places) |

**Recommendation: STABILIZE.**

Gated on `catalog-refactor-shared-json` (supervised pick) + `add-user-role-approval` (supervised pick) + `strategy-library-ddl` (storage). After those land, the entire pending queue (strategy-upload-ui, alex-range-break-* chain, refactor-strategies-page) accelerates. Until then, library work just adds to the catalog-drift debt.

---

## playbooks — **STATUS_QUO**

| metric | value |
|---|---|
| technical debt | LOW |
| business leverage | MEDIUM (teaching/review surface; not in the live execution loop) |
| operational confusion | LOW |
| duplication risk | LOW |

**Recommendation: STATUS_QUO.**

Active and shipping (10 completed in 30d, velocity 0.56). The 4 phase-parallel chart bugs (stacked axis, session-only, ET tz, QA guardrail) are real but low-urgency; they'll get addressed as the chart invariants memory note evolves. Scaffold work (`_modules/scanner`) is fresh per 5/15 daily writeup — let it bake.

---

## pm_system — **STATUS_QUO**

| metric | value |
|---|---|
| technical debt | LOW |
| business leverage | MEDIUM (project-management infrastructure) |
| operational confusion | LOW |
| duplication risk | LOW |

**Recommendation: STATUS_QUO.**

7 open, 7 completed in 30d (velocity 0.50). The nightly task is firing (verified through 5/15). Remaining work (silent failure detection, todos auto-update, visual progress) is incremental.

---

## daily_stop — **STATUS_QUO**

1 open task (`mgp-daily-stop-pipe`, P1, claude, EXECUTION_READY MEDIUM rev). Tool is verified live. Just needs MGP to read the user_settings endpoint.

**Recommendation: STATUS_QUO.**

---

## live_classifier — **STATUS_QUO**

2 open tasks (both roberto-owned). Page shipped 5/5; refresh cron pending. Not on the critical path.

**Recommendation: STATUS_QUO.**

---

## risk_hud — **ARCHIVE candidate** (pending dedup decision)

| metric | value |
|---|---|
| technical debt | HIGH (mock-data page, 5 build tasks blocked on decision) |
| business leverage | UNCERTAIN (acct-v2 may subsume) |
| operational confusion | HIGH |
| duplication risk | HIGHEST in project (risk-bump bug is symptom of dual P&L sources) |

**Recommendation: ARCHIVE pending architecture decision.**

The 5 risk-hud-* build tasks should not ship until `risk-hud-accountability-hud-dedup-review` (P1, user) returns a verdict. If acct-v2 subsumes the use case (likely), archive all 5 + the page itself. If risk_hud stays a separate cockpit, unblock the cluster + execute. **Either way, the decision is the leverage — not the build work.**

---

## mtf_vp_scanner — **STATUS_QUO**

1 open task (`mtf-vp-tradingview-handler`, P1, claude). Path A shipped 4/28 per memory. Path B (this task) is the TradingView ticket-in entry point.

**Recommendation: STATUS_QUO.**

---

## monthly_reports — **STATUS_QUO**

1 open task (`monthly-per-trader-view`, P2, claude, blocked on multi-trader-scorecards). Active and clean.

**Recommendation: STATUS_QUO.**

---

## alerts-hotkeys-etc — **ARCHIVE**

| metric | value |
|---|---|
| business leverage | LOW (1 design task, no completed work) |
| operational confusion | LOW (just 1 task) |

**Recommendation: ARCHIVE the property.**

Per Pass 9 ARCH-2, `hot-key-managment-sell-of-postion` rolls into `overtrading-detect`. With that task closed, the property has no open work. Remove from /system/ hub until a real cluster of hot-key tasks emerges.

---

## Cross-property summary

| Property | Recommendation | Open P0/P1 | Velocity |
|---|---|---|---|
| foundation | **CORE_INFRASTRUCTURE** | 11 | 0.35 |
| scorecards | **STABILIZE** | 16 | 0.40 |
| accountability_hud | **STABILIZE** | 8 | 0.30 |
| mgp | STATUS_QUO (with watch) | 4 | 0.66 |
| param_surface | STABILIZE → AUTOMATE | 3 | 0.25 |
| strategy_library | **STABILIZE** | 4 | 0.27 |
| playbooks | STATUS_QUO | 2 | 0.56 |
| pm_system | STATUS_QUO | 6 | 0.50 |
| daily_stop | STATUS_QUO | 1 | 0.75 |
| live_classifier | STATUS_QUO | 1 | 0.50 |
| risk_hud | **ARCHIVE** (pending decision) | 4 | 0.00 |
| mtf_vp_scanner | STATUS_QUO | 1 | 0.50 |
| monthly_reports | STATUS_QUO | 0 | 0.75 |
| alerts-hotkeys-etc | **ARCHIVE** | 0 | n/a |

**4 STABILIZE / 2 ARCHIVE / 1 CORE_INFRASTRUCTURE / 5 STATUS_QUO + 1 AUTOMATE-after.**

The system-level picture: **foundation needs investment, scorecards needs consolidation, risk_hud needs a decision, everything else is healthy.**
