refactor: consolidate smart routing into model specs (#75)

compute/model-specs/cohere.json

@@ -5,12 +5,25 @@
       "id": "north-mini-code",
       "displayName": "North Mini Code",
       "family": "north",
-      "match": { "exact": ["north-mini-code"], "patterns": ["north-mini-code*"] },
+      "match": {
+        "exact": [
+          "north-mini-code"
+        ],
+        "patterns": [
+          "north-mini-code*"
+        ]
+      },
       "spec": {
         "contextWindow": 256000,
         "maxOutputTokens": 64000,
-        "capabilities": ["chat", "code", "tool_use"],
-        "serviceType": ["chat"],
+        "capabilities": [
+          "chat",
+          "code",
+          "tool_use"
+        ],
+        "serviceType": [
+          "chat"
+        ],
         "supportsReasoning": true,
         "description": "Cohere North 系列首发编程模型"
       }
@@ -19,12 +32,29 @@
       "id": "command-a-03-2025",
       "displayName": "Command A",
       "family": "command-a",
-      "match": { "exact": ["command-a-03-2025"], "patterns": ["command-a*"] },
+      "match": {
+        "exact": [
+          "command-a-03-2025"
+        ],
+        "patterns": [
+          "command-a*"
+        ]
+      },
       "spec": {
         "contextWindow": 256000,
         "maxOutputTokens": 8000,
-        "capabilities": ["chat", "reasoning", "code", "vision", "tool_use", "rag", "long_context"],
-        "serviceType": ["chat"],
+        "capabilities": [
+          "chat",
+          "reasoning",
+          "code",
+          "vision",
+          "tool_use",
+          "rag",
+          "long_context"
+        ],
+        "serviceType": [
+          "chat"
+        ],
         "defaultTemperature": 0.3,
         "description": "Cohere 旗舰对话模型，256K 上下文，擅长 RAG 和工具调用"
       }
@@ -33,12 +63,27 @@
       "id": "command-r7b-12-2024",
       "displayName": "Command R7B",
       "family": "command-r7b",
-      "match": { "exact": ["command-r7b-12-2024"], "patterns": ["command-r7b*"] },
+      "match": {
+        "exact": [
+          "command-r7b-12-2024"
+        ],
+        "patterns": [
+          "command-r7b*"
+        ]
+      },
       "spec": {
         "contextWindow": 128000,
         "maxOutputTokens": 4000,
-        "capabilities": ["chat", "reasoning", "tool_use", "rag", "fast"],
-        "serviceType": ["fast"],
+        "capabilities": [
+          "chat",
+          "reasoning",
+          "tool_use",
+          "rag",
+          "fast"
+        ],
+        "serviceType": [
+          "fast"
+        ],
         "defaultTemperature": 0.3,
         "description": "Cohere 小型高速对话模型，适合高吞吐、低延迟场景"
       }
@@ -47,11 +92,23 @@
       "id": "embed-v4.0",
       "displayName": "Embed V4",
       "family": "embed-v4",
-      "match": { "exact": ["embed-v4.0"], "patterns": ["embed-v4*"] },
+      "match": {
+        "exact": [
+          "embed-v4.0"
+        ],
+        "patterns": [
+          "embed-v4*"
+        ]
+      },
       "spec": {
         "contextWindow": 128000,
-        "capabilities": ["text_embedding", "multilingual"],
-        "serviceType": ["embedding"],
+        "capabilities": [
+          "text_embedding",
+          "multilingual"
+        ],
+        "serviceType": [
+          "embedding"
+        ],
         "description": "Cohere 最新 Embedding 模型，128K 上下文，多语言支持"
       }
     },
@@ -59,11 +116,23 @@
       "id": "rerank-v3.5",
       "displayName": "Rerank V3.5",
       "family": "rerank-v3",
-      "match": { "exact": ["rerank-v3.5"], "patterns": ["rerank-v3*"] },
+      "match": {
+        "exact": [
+          "rerank-v3.5"
+        ],
+        "patterns": [
+          "rerank-v3*"
+        ]
+      },
       "spec": {
         "contextWindow": 4096,
-        "capabilities": ["rerank", "semantic_reranking"],
-        "serviceType": ["rerank"],
+        "capabilities": [
+          "rerank",
+          "semantic_reranking"
+        ],
+        "serviceType": [
+          "rerank"
+        ],
         "description": "Cohere 语义重排序模型，用于检索结果精排"
       }
     }