ryun6249 · ryun6249 · May 19, 2026
diff --git a/app/web/app.js b/app/web/app.js
@@ -3336,6 +3336,7 @@ const UI_LANGUAGE_COPY = {
         momentum: "모멘텀",
         low_volatility: "저변동성",
         liquidity: "유동성",
+        drawdown_resilience: "낙폭 회복",
       },
     },
   },
@@ -3444,6 +3445,7 @@ const UI_LANGUAGE_COPY = {
         momentum: "Momentum",
         low_volatility: "Low Volatility",
         liquidity: "Liquidity",
+        drawdown_resilience: "Drawdown Resilience",
       },
     },
   },
@@ -12242,7 +12244,7 @@ function quantamentalScoreScreenLimit() {
 
 function quantamentalScoreScreenMetric() {
   const raw = String(els.quantamentalScoreMetric?.value || "composite");
-  return ["composite", "value", "quality", "growth", "momentum", "low_volatility", "liquidity"].includes(raw) ? raw : "composite";
+  return ["composite", "value", "quality", "growth", "momentum", "low_volatility", "liquidity", "drawdown_resilience"].includes(raw) ? raw : "composite";
 }
 
 function quantamentalScoreMetricLabel(scoreKey) {

diff --git a/app/web/index.html b/app/web/index.html
@@ -1299,7 +1299,7 @@ <h3>Quantamental</h3>
               <button type="button" id="quantamentalScoreScreenRun" class="linkish" data-testid="quantamental-score-screen-run">실행</button>
             </div>
             <div class="decision-form quantamental-score-screen-form">
-              <label><span>점수 항목</span><select id="quantamentalScoreMetric" aria-label="Quantamental score screen metric"><option value="composite" selected>종합</option><option value="value">가치</option><option value="quality">퀄리티</option><option value="growth">성장</option><option value="momentum">모멘텀</option><option value="low_volatility">저변동성</option><option value="liquidity">유동성</option></select></label>
+                <label><span>점수 항목</span><select id="quantamentalScoreMetric" aria-label="Quantamental score screen metric"><option value="composite" selected>종합</option><option value="value">가치</option><option value="quality">퀄리티</option><option value="growth">성장</option><option value="momentum">모멘텀</option><option value="low_volatility">저변동성</option><option value="liquidity">유동성</option><option value="drawdown_resilience">낙폭 회복</option></select></label>
               <label><span>최소 점수</span><input id="quantamentalScoreThreshold" type="number" min="0" max="100" step="1" value="70" aria-label="Quantamental minimum score" /></label>
               <label><span>결과 수</span><select id="quantamentalScoreScreenLimit" aria-label="Quantamental score screen limit"><option value="10">10</option><option value="20" selected>20</option><option value="50">50</option></select></label>
             </div>
@@ -1613,7 +1613,7 @@ <h2 id="symbolPickerTitle">심볼 찾기</h2>
   <script src="modules/forecast-ui.js?v=20260514-domain-modules"></script>
   <script src="modules/quant-ui.js?v=20260514-domain-modules"></script>
   <script src="modules/ai-portfolio-ui.js?v=20260514-domain-modules"></script>
-  <script src="modules/quantamental-ui.js?v=20260519-quantamental-v13"></script>
-  <script src="app.js?v=20260519-continuous-enhancement-v3"></script>
+  <script src="modules/quantamental-ui.js?v=20260519-quantamental-v14"></script>
+  <script src="app.js?v=20260519-continuous-enhancement-v4"></script>
 </body>
 </html>
diff --git a/app/web/modules/quantamental-ui.js b/app/web/modules/quantamental-ui.js
@@ -1,6 +1,7 @@
 (function initQuantamentalUi(global) {
   const QUALITY_ADJUSTED_MOMENTUM_ID = "quality_adjusted_momentum_v1";
   const VOLATILITY_ADJUSTED_BREAKOUT_ID = "volatility_adjusted_breakout_v1";
+  const DRAWDOWN_RECOVERY_RESILIENCE_ID = "drawdown_recovery_resilience_v1";
 
   function escapeHtml(value) {
     return String(value ?? "")
@@ -73,6 +74,9 @@
       vabScore: "VAB Score",
       vabClass: "VAB Class",
       vabNotInComposite: "Volatility-adjusted breakout is a secondary quant diagnostic and is not used in the composite score.",
+      drsScore: "DRS Score",
+      drsClass: "DRS Class",
+      drsNotInComposite: "Drawdown recovery resilience is a secondary quant diagnostic and is not used in the composite score.",
       maxDrawdown: "Max Drawdown",
       latestFiling: "Latest Filing",
       revenue: "Revenue",
@@ -193,6 +197,7 @@
         momentum: "Momentum",
         lowVolatility: "Low Volatility",
         liquidity: "Liquidity",
+        drawdownResilience: "Drawdown Resilience",
       },
       chart: {
         priceTitle: "Price + SMA",
@@ -288,6 +293,9 @@
       vabScore: "VAB 점수",
       vabClass: "VAB 분류",
       vabNotInComposite: "변동성 조정 돌파는 보조 퀀트 진단 지표이며 복합 점수에는 반영하지 않습니다.",
+      drsScore: "DRS 점수",
+      drsClass: "DRS 분류",
+      drsNotInComposite: "낙폭 회복 탄력성은 보조 퀀트 진단 지표이며 복합 점수에는 반영하지 않습니다.",
       maxDrawdown: "최대 낙폭",
       latestFiling: "최근 공시",
       revenue: "매출",
@@ -408,6 +416,7 @@
         momentum: "모멘텀",
         lowVolatility: "저변동성",
         liquidity: "유동성",
+        drawdownResilience: "낙폭 회복",
       },
       chart: {
         priceTitle: "가격 + SMA",
@@ -567,8 +576,9 @@
     const key = String(value || "").toLowerCase();
     if (key.includes("no_confirmed")) return "warn";
     if (key.includes("confirmed")) return "ok";
+    if (key.includes("resilient")) return "ok";
     if (key.includes("strong") || key.includes("constructive")) return "ok";
-    if (key.includes("weak")) return "fail";
+    if (key.includes("weak") || key.includes("fragile")) return "fail";
     if (key.includes("mixed") || key.includes("insufficient")) return "warn";
     return "neutral";
   }
@@ -605,9 +615,11 @@
     const algorithm = metrics.algorithm || {};
     const algorithms = metrics.algorithms || {};
     const breakout = algorithms.volatility_adjusted_breakout || metrics.volatility_adjusted_breakout || {};
-    if (!algorithm.algorithm_id && !breakout.algorithm_id) return "";
+    const resilience = algorithms.drawdown_recovery_resilience || metrics.drawdown_recovery_resilience || {};
+    if (!algorithm.algorithm_id && !breakout.algorithm_id && !resilience.algorithm_id) return "";
     const algorithmId = algorithm.algorithm_id || QUALITY_ADJUSTED_MOMENTUM_ID;
     const breakoutId = breakout.algorithm_id || VOLATILITY_ADJUSTED_BREAKOUT_ID;
+    const resilienceId = resilience.algorithm_id || DRAWDOWN_RECOVERY_RESILIENCE_ID;
     return `
       ${algorithm.algorithm_id ? `
         <div class="decision-summary ${escapeHtml(algorithmStatusClass(algorithm.classification))}" data-testid="quantamental-quant-algorithm">
@@ -621,6 +633,12 @@
           <br /><span class="muted">${escapeHtml(cpy.vabNotInComposite)}</span>
         </div>
       ` : ""}
+      ${resilience.algorithm_id ? `
+        <div class="decision-summary ${escapeHtml(algorithmStatusClass(resilience.classification))}" data-testid="quantamental-drawdown-resilience-algorithm">
+          ${escapeHtml(resilienceId)} / ${escapeHtml(cpy.drsScore)} ${escapeHtml(fmt(resilience.drawdown_recovery_resilience_score))} / ${escapeHtml(cpy.drsClass)} ${escapeHtml(resilience.classification || cpy.unavailable)}
+          <br /><span class="muted">${escapeHtml(cpy.drsNotInComposite)}</span>
+        </div>
+      ` : ""}
     `;
   }
 
@@ -678,6 +696,7 @@
     ];
     const algorithm = qMetrics?.algorithm || {};
     const breakout = qMetrics?.algorithms?.volatility_adjusted_breakout || {};
+    const resilience = qMetrics?.algorithms?.drawdown_recovery_resilience || {};
     return `
       <div data-testid="quantamental-overview-tab">
         <div class="quantamental-overview-brief">
@@ -690,6 +709,8 @@
             ${metric(cpy.qamClass, algorithm.classification || "-", algorithmStatusClass(algorithm.classification))}
             ${metric(cpy.vabScore, fmt(breakout.volatility_adjusted_breakout_score), scoreClass(breakout.volatility_adjusted_breakout_score))}
             ${metric(cpy.vabClass, breakout.classification || "-", algorithmStatusClass(breakout.classification))}
+            ${metric(cpy.drsScore, fmt(resilience.drawdown_recovery_resilience_score), scoreClass(resilience.drawdown_recovery_resilience_score))}
+            ${metric(cpy.drsClass, resilience.classification || "-", algorithmStatusClass(resilience.classification))}
             ${metric(cpy.maxDrawdown, fmtPct(qMetrics?.drawdown?.max_drawdown), "warn")}
             ${metric(cpy.latestFiling, latestStatement?.date || "-", statusClass(freshness?.sections?.fundamentals?.status))}
             ${metric(cpy.revenue, compact(latestStatement?.revenue), "neutral")}
@@ -1023,6 +1044,7 @@
       momentum: labels.momentum,
       low_volatility: labels.lowVolatility,
       liquidity: labels.liquidity,
+      drawdown_resilience: labels.drawdownResilience,
     }[String(scoreKey || "composite")] || copy().composite;
   }
 

diff --git a/core/schemas/quantamental.py b/core/schemas/quantamental.py
@@ -11,7 +11,16 @@
 QuantamentalPeriod = Literal["annual", "quarterly"]
 QuantamentalStyle = Literal["balanced", "quality_growth", "value", "momentum", "defensive"]
 QuantamentalOutputLanguage = Literal["ko", "en"]
-QuantamentalScoreKey = Literal["composite", "value", "quality", "growth", "momentum", "low_volatility", "liquidity"]
+QuantamentalScoreKey = Literal[
+    "composite",
+    "value",
+    "quality",
+    "growth",
+    "momentum",
+    "low_volatility",
+    "liquidity",
+    "drawdown_resilience",
+]
 
 
 class QuantamentalFlexibleModel(BaseModel):

diff --git a/docs/CONTINUOUS_ENHANCEMENT_LOG.md b/docs/CONTINUOUS_ENHANCEMENT_LOG.md
@@ -228,6 +228,129 @@
 - [x] PR summary includes changed files
 - [x] PR summary includes validation result
 
+## 2026-05-19 Continuous Enhancement Run 15:01
+
+### Current Project Summary
+- Project purpose: FastAPI-served local financial research workstation with deterministic market, macro, Quant Lab, Quantamental, ML Forecast, and AI Portfolio workflows. AI is an interpreter over grounded data, not a replacement for deterministic scoring.
+- Main frontend structure: static `app/web/index.html`, `app/web/app.js`, shared CSS, and domain modules under `app/web/modules`; dashboard panels keep `Core / Diagnostics / Operations / All` with `All` as the default.
+- Main backend structure: FastAPI routes under `app/api/routers`, schema contracts under `core/schemas`, and domain services under `pipelines`.
+- Data flow: UI controls call `/api/v1/*`; routers validate requests; services fetch/cache provider data; deterministic engines produce typed payloads; UI quality/range summaries render from returned payloads.
+- AI/LLM flow: Quantamental AI report/Q&A use deterministic engine snapshots and runtime-checked model selection; default remains deterministic guardrail unless an explicit model request is made.
+- Visualization flow: Quantamental overview renders price, return, volatility, drawdown, volume, fundamentals, and annotated algorithm diagnostics with clear axis notes and missing-value handling.
+- Testing flow: Python/pytest-first with `scripts/check_ui_contract.py`, JS syntax checks, targeted pytest, full pytest, live API smoke, Browser checks, and Playwright smoke scripts. There is still no `package.json`/pnpm build surface.
+
+### Current Problems
+- Compatibility: Existing API contracts are stable; new score-screen keys must stay additive.
+- Data consistency: Current data quality, range, and observation summaries work; new diagnostic algorithms must expose required/available observations and avoid composite-score side effects.
+- UI consistency: Existing QAM/VAB algorithm summaries were visible; additional algorithms need the same concise treatment without cluttering normal tabs.
+- Visualization: Overview already explains axes; new diagnostic values should appear as compact KPI/summary rows.
+- AI briefing: AI context already includes QAM/VAB; new deterministic diagnostics must be added to AI used-data/key-change context without letting AI invent signals.
+- Data freshness: Top-right quality badge remains the user-facing status surface; no freshness rules changed in this run.
+- Translation quality: English/Korean labels must preserve ticker, numeric, date, and unit values.
+- Performance: No background polling or new provider calls should be added for the diagnostic.
+- Code structure: Keep the algorithm inside `pipelines/quantamental/quant_engine.py` and expose it through existing adapters.
+- User experience: Score Threshold Screener should allow screening on the new risk-recovery diagnostic while preserving the single-ticker flow.
+
+### Enhancement Plan
+- Priority 1: Add an additive `drawdown_recovery_resilience_v1` Quantamental diagnostic using existing price/volume/risk-adjusted inputs only.
+- Priority 2: Expose the diagnostic through `quant.metrics.algorithms`, component scores, health metadata, AI context, and score-threshold screening.
+- Priority 3: Add UI labels, summary rows, score-screen option, tests, and live Browser verification without changing strategy/order/composite logic.
+
+### Validation Plan
+- Build: no npm/pnpm build exists; run JS syntax and Python compile/static gates.
+- Lint: run ruff on changed Python surfaces.
+- Unit test: run targeted Quantamental engine/API/UI tests and full pytest.
+- Integration test: smoke `/api/v1/quantamental/health`, `/analysis/AAPL`, and score-screen paths.
+- UI test: Browser desktop/mobile plus `scripts/quantamental_ui_smoke.py`.
+- Data quality test: verify top-right quality summary still shows status, basis date, update time, period, observations, missing, and AI basis after analysis.
+- AI hallucination guard test: verify deterministic AI report includes the new algorithm in `key_changes` and remains advisory-only.
+
+### Changes Made
+- Compatibility: Added `drawdown_resilience` as an additive `QuantamentalScoreKey`; existing score keys and default composite screening remain unchanged.
+- Data consistency: Added `drawdown_recovery_resilience_v1` with required/available observations, input provenance, component scores, warnings, and `used_in_composite_score=false`.
+- UI/UX: Added DRS score/class rows to Quantamental overview and score summary, plus a Score Threshold Screener option labeled `Drawdown Resilience` / `낙폭 회복`.
+- Visualization: DRS now appears next to QAM/VAB in the compact overview metric strip and algorithm summary blocks.
+- AI Briefing: Added DRS to `quant_snapshot` and deterministic AI `key_changes` as `drawdown_recovery_algorithm`; AI still interprets deterministic outputs only.
+- Translation: Added concise English/Korean labels; ticker/date/number/unit handling was not changed.
+- Performance: Reused already-loaded price, return, volatility, drawdown, risk-adjusted, and liquidity vectors; no provider call or background polling was added.
+- Safety: No trading/order execution code, strategy entry/exit logic, secrets, `.env`, provider selection, or composite scoring weights were changed.
+
+### 15:01 Validation Results
+
+| Check | Command / Tool | Result | Notes |
+|---|---|---|---|
+| Python syntax | `python -m py_compile pipelines\quantamental\quant_engine.py pipelines\quantamental\service.py pipelines\quantamental\ai_service.py core\schemas\quantamental.py scripts\check_ui_contract.py scripts\quantamental_ui_smoke.py scripts\ai_portfolio_ui_smoke.py` | Passed | Changed Python surfaces compile. |
+| JS syntax | `node --check app\web\modules\quantamental-ui.js` and `node --check app\web\app.js` | Passed | Static UI JavaScript syntax. |
+| Lint | `python -m ruff check pipelines\quantamental\quant_engine.py pipelines\quantamental\service.py pipelines\quantamental\ai_service.py core\schemas\quantamental.py tests\test_quantamental_api.py tests\test_quantamental_engines.py` | Passed | Changed Python implementation/tests pass lint. |
+| UI contract | `python scripts\check_ui_contract.py` | Passed | v14 Quantamental bundle, DRS markers, and UI contracts present. |
+| Targeted tests | `python -m pytest tests\test_quantamental_engines.py tests\test_quantamental_api.py tests\test_ui_modules.py tests\test_ui_routing_contract.py -q` | Passed | `81 passed, 4 subtests passed`. |
+| Full regression | `python -m pytest -q` | Passed | `693 passed, 9 subtests passed`. |
+| Live health/API | `GET /api/v1/quantamental/health`; `GET /api/v1/quantamental/analysis/AAPL?include_ai=true&use_llm=false&lookback=252&output_language=en` on `127.0.0.1:8000` | Passed | Health lists `drawdown_recovery_resilience_v1`; AAPL returned DRS score and AI `drawdown_recovery_algorithm`. |
+| Browser desktop UI | Browser at `http://127.0.0.1:8000/ui/?range=1Y#quantamental` | Passed | `panelView=all`, range `1Y`, quality badge populated, DRS visible, Drawdown Resilience screen returned 10 rows, no console errors or horizontal overflow. |
+| Browser mobile UI | Browser viewport `390x900` | Passed | All view, top quality summary, DRS score option, and no horizontal overflow confirmed. |
+| Quantamental browser smoke | `python scripts\quantamental_ui_smoke.py --base-url http://127.0.0.1:8000 --output reports\quantamental_ui_smoke_continuous_20260519_1501.json` | Passed | Required ticker set, invalid ticker, GLOBAL resolver, Top 5, Drawdown Resilience score screen, overview axes, AI/Q&A, and audit smoke passed. |
+| Cross-tab browser smoke | `python scripts\ai_portfolio_ui_smoke.py --base-url http://127.0.0.1:8000 --timeout-s 180 --output reports\ai_portfolio_ui_smoke_continuous_20260519_1501.json` | Retried | First run timed out on Macro series search after 180s; console errors were empty. |
+| Cross-tab browser smoke retry | `python scripts\ai_portfolio_ui_smoke.py --base-url http://127.0.0.1:8000 --timeout-s 240 --output reports\ai_portfolio_ui_smoke_continuous_20260519_1501_retry.json` | Passed | Versioned scripts, dashboard tab matrix, Quantamental language/top5/score screen, and action smoke passed. |
+| npm/pnpm build/lint/test | Not run | Excluded | Repo has no `package.json`, `pnpm-lock.yaml`, or frontend build manifest. |
+
+### 15:01 Completion Checklist
+
+#### Compatibility
+- [x] Existing features still work
+- [x] Existing API contracts are not broken
+- [x] Existing UI flow is preserved
+- [x] No unauthorized strategy logic change
+- [x] No secret or env file exposure
+
+#### Data
+- [x] Date range selection works
+- [x] KPI/chart/table use the same selected period in the checked Quantamental flow
+- [x] Data source and basis date are displayed
+- [x] Missing data is handled
+- [x] Data quality summary is visible at top-right
+- [x] Cache/fresh data distinction is clear
+
+#### UI
+- [x] Default view is All
+- [x] Core/Diagnostics/Operations filters still exist
+- [x] Font sizes are readable in the checked Quantamental surface
+- [x] Layout spacing is consistent in checked desktop/mobile surfaces
+- [x] Cards/tables/charts are aligned
+- [x] Mobile layout is acceptable
+- [x] Loading state exists
+- [x] Empty state exists
+- [x] Error state exists
+
+#### Visualization
+- [x] Chart titles are meaningful
+- [x] Axis labels are readable
+- [x] Tooltips/legends remain useful
+- [x] Period selection updates checked Quantamental results
+- [x] No chart overflow or label collision observed in Browser checks
+
+#### AI Briefing
+- [x] Gemma/Qwen availability remains runtime-checked
+- [x] Model selection is not fake
+- [x] AI output includes used data period
+- [x] AI output includes basis/source/observation count
+- [x] AI does not invent unsupported numbers
+- [x] Unverified facts are marked unavailable by existing guardrails
+- [x] Translation preserves numbers/dates/units in tested API/UI paths
+
+#### Validation
+- [x] Lint executed or reason documented
+- [x] Build executed or reason documented
+- [x] Tests executed or reason documented
+- [x] UI validation executed or reason documented
+- [x] Data validation executed or reason documented
+- [x] AI briefing validation executed or reason documented
+
+#### Documentation
+- [x] docs/CONTINUOUS_ENHANCEMENT_LOG.md updated
+- [x] README updated if needed
+- [x] PR summary includes changed files
+- [x] PR summary includes validation result
+
 ## 2026-05-19 Continuous Enhancement Run 09:20 Final
 
 - Branch: `automation/continuous-enhancement-20260519-0920`.