@dotobokuri/fleet-console 1.53.0 → 1.54.0

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

@@ -17504,92 +17504,168 @@ 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",
@@ -17726,7 +17802,7 @@ 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",
@@ -17785,94 +17861,6 @@ 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"
         }
       ]
     },
@@ -18405,7 +18393,8 @@ var ANTHROPIC_EFFORT_RUNGS = /* @__PURE__ */ new Set([
   "medium",
   "high",
   "xhigh",
-  "max"
+  "max",
+  "ultra"
 ]);
 function freezeGatewayModelEffort(effort) {
   if (!effort?.supported) return UNSUPPORTED_GATEWAY_MODEL_EFFORT;
@@ -19417,7 +19406,7 @@ function runVendorQuery(input) {
 }
 
 // ../../packages/core-agent/src/claude/sdk.ts
-var NATIVE_MODEL_ALIASES = /* @__PURE__ */ new Set(["sonnet", "opus", "haiku", "fable"]);
+var NATIVE_MODEL_ALIASES = /* @__PURE__ */ new Set(["sonnet", "opus", "haiku", "fable", "fable[1m]"]);
 async function createClaudeGatewaySdk(options) {
   const baseUrl = normalizeBaseUrl(options.baseUrl);
   const accepted = resolveModels(options.models);