@dotobokuri/fleet-console 1.53.0 → 1.54.1

package/dist/fleet-plugins/quota/routes.mjs

@@ -17826,92 +17826,168 @@ var benchmarks_default = {
     "gpt-5.6-sol": {
       source: "cursorbench",
       rungs: {
-        low: { score: 52.6, tokensPerTask: 5104, stepsPerTask: 19 },
-        medium: { score: 60, tokensPerTask: 9747, stepsPerTask: 27 },
-        high: { score: 63.5, tokensPerTask: 13867, stepsPerTask: 32 },
-        xhigh: { score: 64.5, tokensPerTask: 19699, stepsPerTask: 38 },
-        max: { score: 67.2, tokensPerTask: 28320, stepsPerTask: 48 }
+        low: {
+          score: 52.6,
+          tokensPerTask: 5104,
+          stepsPerTask: 19
+        },
+        medium: {
+          score: 60,
+          tokensPerTask: 9747,
+          stepsPerTask: 27
+        },
+        high: {
+          score: 63.5,
+          tokensPerTask: 13867,
+          stepsPerTask: 32
+        },
+        xhigh: {
+          score: 64.5,
+          tokensPerTask: 19699,
+          stepsPerTask: 38
+        },
+        max: {
+          score: 67.2,
+          tokensPerTask: 28320,
+          stepsPerTask: 48
+        }
       }
     },
     "gpt-5.6-terra": {
       source: "cursorbench",
       rungs: {
-        low: { score: 46.9, tokensPerTask: 5312, stepsPerTask: 19 },
-        medium: { score: 50.3, tokensPerTask: 6222, stepsPerTask: 20 },
-        high: { score: 54.2, tokensPerTask: 9468, stepsPerTask: 23 },
-        xhigh: { score: 59.2, tokensPerTask: 16089, stepsPerTask: 29 },
-        max: { score: 64.9, tokensPerTask: 32969, stepsPerTask: 47 }
+        low: {
+          score: 46.9,
+          tokensPerTask: 5312,
+          stepsPerTask: 19
+        },
+        medium: {
+          score: 50.3,
+          tokensPerTask: 6222,
+          stepsPerTask: 20
+        },
+        high: {
+          score: 54.2,
+          tokensPerTask: 9468,
+          stepsPerTask: 23
+        },
+        xhigh: {
+          score: 59.2,
+          tokensPerTask: 16089,
+          stepsPerTask: 29
+        },
+        max: {
+          score: 64.9,
+          tokensPerTask: 32969,
+          stepsPerTask: 47
+        }
       }
     },
     "gpt-5.6-luna": {
       source: "cursorbench",
       rungs: {
-        low: { score: 37.6, tokensPerTask: 3209, stepsPerTask: 17 },
-        medium: { score: 47.7, tokensPerTask: 7095, stepsPerTask: 28 },
-        high: { score: 56.8, tokensPerTask: 15141, stepsPerTask: 40 },
-        xhigh: { score: 57.7, tokensPerTask: 22480, stepsPerTask: 48 },
-        max: { score: 61.1, tokensPerTask: 87973, stepsPerTask: 61 }
+        low: {
+          score: 37.6,
+          tokensPerTask: 3209,
+          stepsPerTask: 17
+        },
+        medium: {
+          score: 47.7,
+          tokensPerTask: 7095,
+          stepsPerTask: 28
+        },
+        high: {
+          score: 56.8,
+          tokensPerTask: 15141,
+          stepsPerTask: 40
+        },
+        xhigh: {
+          score: 57.7,
+          tokensPerTask: 22480,
+          stepsPerTask: 48
+        },
+        max: {
+          score: 61.1,
+          tokensPerTask: 87973,
+          stepsPerTask: 61
+        }
       }
     },
     "grok-4.5": {
       source: "cursorbench",
       rungs: {
-        low: { score: 63.5, tokensPerTask: 15841, stepsPerTask: 31 },
-        medium: { score: 65.4, tokensPerTask: 18914, stepsPerTask: 34 },
-        high: { score: 66.7, tokensPerTask: 19521, stepsPerTask: 33 }
+        low: {
+          score: 63.5,
+          tokensPerTask: 15841,
+          stepsPerTask: 31
+        },
+        medium: {
+          score: 65.4,
+          tokensPerTask: 18914,
+          stepsPerTask: 34
+        },
+        high: {
+          score: 66.7,
+          tokensPerTask: 19521,
+          stepsPerTask: 33
+        }
       },
       caveat: "Bench footnote: an earlier snapshot of Cursor's codebase was unintentionally present in this model's training data, inflating scores by an unknown amount; the advantage does not transfer to other repositories."
     },
     "kimi-k3": {
       source: "cursorbench",
       rungs: {
-        low: { score: 50.5, tokensPerTask: 13007, stepsPerTask: 33 },
-        high: { score: 59.7, tokensPerTask: 26846, stepsPerTask: 47 },
-        max: { score: 60.8, tokensPerTask: 38428, stepsPerTask: 57 }
+        low: {
+          score: 50.5,
+          tokensPerTask: 13007,
+          stepsPerTask: 33
+        },
+        high: {
+          score: 59.7,
+          tokensPerTask: 26846,
+          stepsPerTask: 47
+        },
+        max: {
+          score: 60.8,
+          tokensPerTask: 38428,
+          stepsPerTask: 57
+        }
       }
     },
     "composer-2.5": {
       source: "cursorbench",
-      overall: { score: 56.1, tokensPerTask: 14286, stepsPerTask: 33 }
+      overall: {
+        score: 56.1,
+        tokensPerTask: 14286,
+        stepsPerTask: 33
+      }
     },
     "glm-5.2": {
       source: "cursorbench",
       rungs: {
-        high: { score: 51.5, tokensPerTask: 21829, stepsPerTask: 49 },
-        max: { score: 55, tokensPerTask: 35946, stepsPerTask: 58 }
+        high: {
+          score: 51.5,
+          tokensPerTask: 21829,
+          stepsPerTask: 49
+        },
+        max: {
+          score: 55,
+          tokensPerTask: 35946,
+          stepsPerTask: 58
+        }
       },
       caveat: "Measured per reasoning rung, but the gateway serves this model without effort control; which rung the serving path reaches is unknown."
-    },
-    "claude-opus-5": {
-      source: "cursorbench",
-      rungs: {
-        low: { score: 62.8, tokensPerTask: 18529, stepsPerTask: 37 },
-        medium: { score: 64.3, tokensPerTask: 23612, stepsPerTask: 44 },
-        high: { score: 66.7, tokensPerTask: 27932, stepsPerTask: 48 },
-        xhigh: { score: 69.3, tokensPerTask: 54239, stepsPerTask: 72 },
-        max: { score: 70, tokensPerTask: 61838, stepsPerTask: 78 }
-      }
-    },
-    "claude-fable-5": {
-      source: "cursorbench",
-      rungs: {
-        low: { score: 62.1, tokensPerTask: 18182, stepsPerTask: 31 },
-        medium: { score: 65.2, tokensPerTask: 30366, stepsPerTask: 41 },
-        high: { score: 66.5, tokensPerTask: 43747, stepsPerTask: 48 },
-        xhigh: { score: 68.4, tokensPerTask: 64971, stepsPerTask: 56 },
-        max: { score: 70.5, tokensPerTask: 103525, stepsPerTask: 72 }
-      }
     }
   }
 };
 var models_default = {
   version: 1,
-  updatedAt: "2026-08-08T00:00:00Z",
+  updatedAt: "2026-08-11T00:00:00Z",
   providers: {
     codex: {
       name: "Codex",
       defaultModel: "gpt-5.6-sol",
-      source: "codex app-server model/list (codex-cli 0.146.0, 2026-08-01) for model/effort/service-tier; contextWindow=272000 is codex debug models context_window/max_context_window (codex-cli 0.146.0, 2026-08-02)",
+      source: "codex app-server model/list (codex-cli 0.147.0, 2026-08-11 \u2014 ultra advertised for sol/sol-wm/terra, luna stops at max) for model/effort/service-tier; contextWindow=272000 is codex debug models context_window/max_context_window (codex-cli 0.146.0, 2026-08-02)",
       models: [
         {
           modelId: "gpt-5.6-sol",
@@ -18048,7 +18124,7 @@ var models_default = {
     cursor: {
       name: "Cursor",
       defaultModel: "auto",
-      source: "Cursor GetUsableModels + Run conversationCheckpointUpdate.token_details.max_tokens (cli-2026.07.08-0c04a8a; standard and Max Mode observed 2026-08-01, gpt-5.6-sol observed 2026-08-05 on cursor-agent 2026.07.23-e383d2b); quotaScope from DashboardService/GetCurrentPeriodUsage autoBucketModels (2026-08-05)",
+      source: "Cursor GetUsableModels + Run conversationCheckpointUpdate.token_details.max_tokens (cli-2026.07.08-0c04a8a; auto, composer-2.5, and grok-4.5 families observed 2026-08-01); quotaScope from DashboardService/GetCurrentPeriodUsage autoBucketModels (2026-08-05)",
       models: [
         {
           modelId: "auto",
@@ -18107,94 +18183,6 @@ var models_default = {
             ],
             upstreamModelIdTemplate: "cursor-grok-4.5-{effort}-fast"
           }
-        },
-        {
-          modelId: "gpt-5.6-sol",
-          name: "GPT-5.6-Sol",
-          capabilityClass: "flagship",
-          quotaScope: "api",
-          contextWindow: 272e3,
-          effort: {
-            supported: true,
-            levels: [
-              "low",
-              "medium",
-              "high",
-              "xhigh",
-              "max"
-            ],
-            upstreamModelIdTemplate: "gpt-5.6-sol-{effort}"
-          },
-          benchmarkKey: "gpt-5.6-sol"
-        },
-        {
-          modelId: "claude-opus-5",
-          name: "Opus-5",
-          capabilityClass: "flagship",
-          quotaScope: "api",
-          contextWindow: 3e5,
-          effort: {
-            supported: true,
-            levels: [
-              "low",
-              "medium",
-              "high",
-              "xhigh",
-              "max"
-            ],
-            upstreamModelIdTemplate: "claude-opus-5-{effort}",
-            upstreamModelIds: {
-              xhigh: "claude-opus-5-thinking-xhigh",
-              max: "claude-opus-5-thinking-max"
-            }
-          },
-          benchmarkKey: "claude-opus-5"
-        },
-        {
-          modelId: "claude-fable-5",
-          name: "Fable-5",
-          capabilityClass: "flagship",
-          quotaScope: "api",
-          description: "NO ZDR",
-          contextWindow: 3e5,
-          effort: {
-            supported: true,
-            levels: [
-              "low",
-              "medium",
-              "high",
-              "xhigh",
-              "max"
-            ],
-            upstreamModelIdTemplate: "claude-fable-5-{effort}"
-          },
-          benchmarkKey: "claude-fable-5"
-        },
-        {
-          modelId: "kimi-k3-1m",
-          name: "Kimi-K3-1M",
-          capabilityClass: "flagship",
-          quotaScope: "api",
-          providerModelId: "kimi-k3-max",
-          contextWindow: 1048576,
-          cursorMaxMode: true,
-          benchmarkKey: "kimi-k3"
-        },
-        {
-          modelId: "kimi-k3",
-          name: "Kimi-K3",
-          capabilityClass: "flagship",
-          quotaScope: "api",
-          contextWindow: 2e5,
-          effort: {
-            supported: true,
-            levels: [
-              "low",
-              "high"
-            ],
-            upstreamModelIdTemplate: "kimi-k3-{effort}"
-          },
-          benchmarkKey: "kimi-k3"
         }
       ]
     },
@@ -18690,6 +18678,62 @@ function freezeGatewayModelEffort(effort) {
 upstreamModelId(gatewayProviderDefault("codex"));
 new Set(REASONING_EFFORTS);
 new TextEncoder().encode(": keep-alive\n\n");
+var MS_PER_HOUR = 36e5;
+var CADENCE_SESSION_MAX_MS = 20 * MS_PER_HOUR;
+var CADENCE_DAILY_MAX_MS = 3 * 24 * MS_PER_HOUR;
+var CADENCE_WEEKLY_MAX_MS = 20 * 24 * MS_PER_HOUR;
+var MIN_ELAPSED_FRACTION = 0.05;
+var PACE_CRITICAL = 1.5;
+var PACE_ELEVATED = 1.1;
+var USED_CRITICAL_PERCENT = 95;
+var USED_ELEVATED_PERCENT = 80;
+function quotaWindowCadence(durationMs) {
+  if (durationMs <= CADENCE_SESSION_MAX_MS) return "session";
+  if (durationMs <= CADENCE_DAILY_MAX_MS) return "daily";
+  if (durationMs <= CADENCE_WEEKLY_MAX_MS) return "weekly";
+  return "monthly";
+}
+function windowPressure(usedPercent, paceRatio) {
+  if (usedPercent >= USED_CRITICAL_PERCENT || paceRatio !== void 0 && paceRatio >= PACE_CRITICAL) {
+    return "critical";
+  }
+  if (usedPercent >= USED_ELEVATED_PERCENT || paceRatio !== void 0 && paceRatio >= PACE_ELEVATED) {
+    return "elevated";
+  }
+  return "ok";
+}
+function deriveQuotaWindowRisk(window, at) {
+  const durationMs = window.period?.durationMs;
+  if (typeof durationMs !== "number" || !Number.isFinite(durationMs) || durationMs <= 0) {
+    return { pressure: windowPressure(window.usedPercent, void 0) };
+  }
+  const startsAt = window.period?.startsAt ?? (window.resetsAt !== void 0 && window.resetsAt > durationMs ? window.resetsAt - durationMs : void 0);
+  const resetBoundary = window.resetsAt ?? (startsAt !== void 0 ? startsAt + durationMs : void 0);
+  const stale = resetBoundary !== void 0 && at > resetBoundary;
+  let elapsedFraction;
+  let paceRatio;
+  let projectedExhaustionAt;
+  if (startsAt !== void 0 && resetBoundary !== void 0 && !stale && at > startsAt) {
+    const elapsed = Math.min(1, (at - startsAt) / durationMs);
+    elapsedFraction = Math.round(elapsed * 100) / 100;
+    if (elapsed >= MIN_ELAPSED_FRACTION) {
+      const used = Math.min(1, Math.max(0, window.usedPercent / 100));
+      paceRatio = Math.round(used / elapsed * 100) / 100;
+      if (used > 0 && used < 1) {
+        const exhaustionAt = startsAt + Math.round((at - startsAt) / used);
+        if (exhaustionAt < resetBoundary) projectedExhaustionAt = exhaustionAt;
+      }
+    }
+  }
+  return {
+    cadence: quotaWindowCadence(durationMs),
+    ...elapsedFraction !== void 0 ? { elapsedFraction } : {},
+    ...paceRatio !== void 0 ? { paceRatio } : {},
+    ...projectedExhaustionAt !== void 0 ? { projectedExhaustionAt } : {},
+    recoveryHalfLifeMs: Math.round(durationMs / 2),
+    pressure: windowPressure(window.usedPercent, paceRatio)
+  };
+}
 var REQUEST_TIMEOUT_MS = 1e4;
 var MAX_RESPONSE_BYTES = 262144;
 var MAX_WINDOWS = 8;
@@ -19634,6 +19678,14 @@ function createQuotaService(deps) {
     inFlight.set(id, task);
     return task;
   }
+  function withRisk(provider) {
+    if (!provider.windows || provider.windows.length === 0) return provider;
+    const at = typeof provider.fetchedAt === "number" && Number.isFinite(provider.fetchedAt) ? provider.fetchedAt : now();
+    return {
+      ...provider,
+      windows: provider.windows.map((window) => ({ ...window, risk: deriveQuotaWindowRisk(window, at) }))
+    };
+  }
   return {
     async getSummary(options = {}) {
       const [claude, codex, cursor, kimi, opencode] = await Promise.all([
@@ -19643,7 +19695,15 @@ function createQuotaService(deps) {
         load("kimi", options.force === true || options.forceProvider === "kimi"),
         load("opencode", options.force === true || options.forceProvider === "opencode")
       ]);
-      return { providers: { claude, codex, cursor, kimi, opencode } };
+      return {
+        providers: {
+          claude: withRisk(claude),
+          codex: withRisk(codex),
+          cursor: withRisk(cursor),
+          kimi: withRisk(kimi),
+          opencode: withRisk(opencode)
+        }
+      };
     }
   };
 }