feat(manifests): update schema, mapping, and existing model manifests

Update model schema with new fields, add mapping configuration, update 24 existing model manifests with enhanced metadata, and remove deprecated llama-4-maverick manifest.

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

Author: ericyangpan

docs/MANIFEST_I18N.md

@@ -77,7 +77,7 @@ export default function MyComponent() {
 import { getTranslations } from 'next-intl/server'
 
 export default async function Page() {
-  const t = await getTranslations('home')
+  const tPage = await getTranslations('home')
   return <h1>{t('title')}</h1>
 }
 ```
@@ -282,7 +282,7 @@ Include translations for all supported locales in manifest files:
 |------|---------|
 | `src/i18n/config.ts` | Locale configuration |
 | `src/i18n/navigation.ts` | Localized Link component |
-| `src/i18n/lib-core.mjs` | Reference resolution |
+| `src/i18n/lib-core.ts` | Reference resolution |
 | `src/i18n/request.ts` | Request config for next-intl |
 | `src/lib/manifest-i18n.ts` | Manifest translation layer |
 | `translations/{locale}/` | UI translation files |

docs/SCHEMA-ALIGNMENT.md

@@ -135,15 +135,13 @@ export interface ManifestCommunityUrls {
   discord: string | null
   reddit: string | null
   blog: string | null
-  [key: string]: string | null
 }
 
 // manifests/$schemas/ref/platform-urls.schema.json → ManifestPlatformUrls
 export interface ManifestPlatformUrls {
   huggingface: string | null
   artificialAnalysis: string | null
   openrouter: string | null
-  [key: string]: string | null
 }
 ```
 
@@ -177,7 +175,6 @@ export interface ManifestIDE extends ManifestBaseApp {}
 // manifests/$schemas/extension.schema.json → ManifestExtension
 export interface ManifestExtension extends ManifestBaseProduct {
   supportedIdes: ManifestIDESupport[]
-  platforms?: ManifestPlatformElement[]
 }
 ```
 
@@ -191,8 +188,8 @@ export interface ManifestModel extends ManifestVendorEntity {
   maxOutput: number
   tokenPricing: ManifestTokenPricing
   releaseDate: string | null
-  inputModalities: ('image' | 'text' | 'file')[]
-  capabilities: ('function-calling' | 'tool-choice' | 'structured-outputs' | 'reasoning')[]
+  inputModalities: ModelInputModality[]
+  capabilities: ModelCapability[]
   benchmarks: ManifestBenchmarks
   platformUrls: ManifestPlatformUrls
 }
@@ -207,7 +204,6 @@ export interface ManifestProvider extends ManifestVendorEntity {
 
 // manifests/$schemas/vendor.schema.json → ManifestVendor
 export interface ManifestVendor extends ManifestEntity {
-  docsUrl?: string | null
   communityUrls: ManifestCommunityUrls
 }
 ```
@@ -248,15 +244,9 @@ export interface ManifestCollectionItem {
 
 ## Type Guards
 
-The system includes type guard functions for runtime type checking:
-
-```typescript
-export function isManifestEntity(obj: unknown): obj is ManifestEntity
-export function isManifestVendorEntity(obj: unknown): obj is ManifestVendorEntity
-export function isManifestBaseProduct(obj: unknown): obj is ManifestBaseProduct
-export function isManifestModel(obj: unknown): obj is ManifestModel
-export function isManifestProvider(obj: unknown): obj is ManifestProvider
-```
+Type guards are intentionally **not** defined in `src/types/manifests.ts` to keep it type-only.
+If runtime guards are needed, define them in a separate module under `src/lib/` (or `src/types/guards/`)
+so client components don't accidentally pull in runtime code when importing types.
 
 ---
 

manifests/$schemas/model.schema.json

@@ -45,14 +45,24 @@
         "releaseDate": {
           "type": ["string", "null"],
           "format": "date",
-          "description": "Model release date in ISO 8601 format (YYYY-MM-DD), null if release date is unknown"
+          "description": "Model release date (ISO 8601 date, YYYY-MM-DD), null if unknown"
+        },
+        "lifecycle": {
+          "type": "string",
+          "enum": ["latest", "maintained", "deprecated"],
+          "description": "Lifecycle stage: latest (newly released), maintained (still available for use), deprecated (no longer available)"
+        },
+        "knowledgeCutoff": {
+          "type": ["string", "null"],
+          "pattern": "^\\d{4}-(0[1-9]|1[0-2])(?:-(0[1-9]|[12]\\d|3[01]))?$",
+          "description": "Knowledge cutoff date when the model's training data ended (ISO 8601, YYYY-MM or YYYY-MM-DD), null if unknown"
         },
         "inputModalities": {
           "type": "array",
           "description": "Input modalities supported by the model",
           "items": {
             "type": "string",
-            "enum": ["image", "text", "file"]
+            "enum": ["text", "image", "pdf", "audio", "video"]
           },
           "uniqueItems": true,
           "minItems": 1
@@ -121,6 +131,8 @@
         "maxOutput",
         "tokenPricing",
         "releaseDate",
+        "lifecycle",
+        "knowledgeCutoff",
         "inputModalities",
         "capabilities",
         "benchmarks",

manifests/mapping.json

@@ -0,0 +1,27 @@
+{
+  "vendors": {
+    "Meta": "llama",
+    "Moonshot": "moonshotai",
+    "Z.ai": "zai"
+  },
+  "models": {
+    "claude-opus-4": "claude-opus-4-0",
+    "claude-sonnet-4": "claude-sonnet-4-0",
+    "glm-4-6": "glm-4.6",
+    "glm-4-6v": "glm-4.6v",
+    "glm-4-7": "glm-4.7",
+    "gpt-4-1": "gpt-4.1",
+    "gpt-5-1": "gpt-5.1",
+    "gpt-5-2": "gpt-5.2",
+    "gpt-5-1-codex": "gpt-5.1-codex",
+    "qwen3-coder-30b-a3b": "qwen3-coder-30b-a3b-instruct",
+    "qwen3-coder-480b-a35b": "qwen3-coder-480b-a35b-instruct",
+    "minimax-m2": "MiniMax-M2",
+    "minimax-m2-1": "MiniMax-M2.1",
+    "gemini-2-5-flash": "gemini-2.5-flash",
+    "gemini-2-5-pro": "gemini-2.5-pro-preview-05-06",
+    "gemini-3-pro": "gemini-3-pro-preview",
+    "gemini-3-flash": "gemini-3-flash-preview",
+    "kimi-k2-0905": "kimi-k2-0905-preview"
+  }
+}

manifests/models/claude-haiku-4-5.json

@@ -38,20 +38,22 @@
       "description": "Anthropic 最快、最高效的模型，以大型 Claude 模型的一小部分成本和延遲提供接近前沿的智能。在推理、編碼和計算機使用任務方面與 Claude Sonnet 4 的性能相當，Haiku 4.5 為實時和大规模應用帶來前沿級能力。"
     }
   },
-  "verified": false,
+  "verified": true,
   "websiteUrl": "https://www.anthropic.com",
   "docsUrl": "https://docs.anthropic.com/claude/docs/models-overview",
   "vendor": "Anthropic",
   "size": "Unknown",
   "contextWindow": 200000,
-  "maxOutput": 8000,
+  "maxOutput": 64000,
   "tokenPricing": {
     "input": 1,
-    "output": 1,
-    "cache": null
+    "output": 5,
+    "cache": 0.1
   },
-  "releaseDate": null,
-  "inputModalities": ["text"],
+  "releaseDate": "2025-10-15",
+  "lifecycle": "latest",
+  "knowledgeCutoff": "2025-02",
+  "inputModalities": ["text", "image", "pdf"],
   "capabilities": ["function-calling", "tool-choice", "structured-outputs", "reasoning"],
   "benchmarks": {
     "sweBench": null,

manifests/models/claude-opus-4-1.json

@@ -38,21 +38,23 @@
       "description": "2025年8月發布的 Claude Opus 4.1，是 Anthropic 最新的旗艦模型，具備卓越的編碼能力，在 HumanEval+ 上達到 85.2%（首次超越 GPT-4o 的 84.9%），在 SWE-bench Verified 上達到 74.5%，擁有 200K 上下文視窗，適用於複雜的代碼生成任務。"
     }
   },
-  "verified": false,
+  "verified": true,
   "websiteUrl": "https://www.anthropic.com",
   "docsUrl": "https://docs.anthropic.com/claude/docs/models-overview",
   "vendor": "Anthropic",
   "size": "Unknown",
   "contextWindow": 200000,
-  "maxOutput": 8000,
+  "maxOutput": 32000,
   "tokenPricing": {
     "input": 15,
     "output": 75,
-    "cache": null
+    "cache": 1.5
   },
-  "releaseDate": "2025-08-01",
-  "inputModalities": ["text"],
-  "capabilities": ["function-calling", "tool-choice", "structured-outputs"],
+  "releaseDate": "2025-08-05",
+  "lifecycle": "maintained",
+  "knowledgeCutoff": null,
+  "inputModalities": ["text", "image", "pdf"],
+  "capabilities": ["function-calling", "reasoning", "tool-choice", "structured-outputs"],
   "benchmarks": {
     "sweBench": 74.4,
     "terminalBench": 0.631,

manifests/models/claude-opus-4.json

@@ -38,21 +38,23 @@
       "description": "Claude Opus 4 是 Anthropic 推出的 200K 上下文視窗語言模型，專為複雜的代碼生成任務設計。"
     }
   },
-  "verified": false,
+  "verified": true,
   "websiteUrl": "https://www.anthropic.com",
   "docsUrl": "https://docs.anthropic.com/claude/docs/models-overview",
   "vendor": "Anthropic",
   "size": "Unknown",
   "contextWindow": 200000,
-  "maxOutput": 8000,
+  "maxOutput": 32000,
   "tokenPricing": {
     "input": 15,
     "output": 75,
-    "cache": null
+    "cache": 1.5
   },
-  "releaseDate": null,
-  "inputModalities": ["text"],
-  "capabilities": ["function-calling", "tool-choice", "structured-outputs"],
+  "releaseDate": "2025-05-22",
+  "lifecycle": "maintained",
+  "knowledgeCutoff": null,
+  "inputModalities": ["text", "image", "pdf"],
+  "capabilities": ["function-calling", "reasoning", "tool-choice", "structured-outputs"],
   "benchmarks": {
     "sweBench": 67.6,
     "terminalBench": 0.351,

manifests/models/claude-sonnet-4-5.json

@@ -38,21 +38,23 @@
       "description": "Anthropic 迄今為止最先進的 Sonnet 模型，針對真實世界的代理和編碼工作流程進行了優化。在 SWE-bench Verified 等編碼基準測試中表現出色，在系統設計、代碼安全和規範遵循方面有所提升。"
     }
   },
-  "verified": false,
+  "verified": true,
   "websiteUrl": "https://www.anthropic.com",
   "docsUrl": "https://docs.anthropic.com/claude/docs/models-overview",
   "vendor": "Anthropic",
   "size": "Unknown",
   "contextWindow": 200000,
-  "maxOutput": 8000,
+  "maxOutput": 64000,
   "tokenPricing": {
     "input": 3,
-    "output": 3,
-    "cache": null
+    "output": 15,
+    "cache": 0.3
   },
-  "releaseDate": null,
-  "inputModalities": ["text"],
-  "capabilities": ["function-calling", "tool-choice", "structured-outputs"],
+  "releaseDate": "2025-09-29",
+  "lifecycle": "latest",
+  "knowledgeCutoff": "2025-01",
+  "inputModalities": ["text", "image", "pdf"],
+  "capabilities": ["function-calling", "reasoning", "tool-choice", "structured-outputs"],
   "benchmarks": {
     "sweBench": 70.6,
     "terminalBench": 0.42,

manifests/models/claude-sonnet-4.json

@@ -38,20 +38,22 @@
       "description": "顯著增強了其前身 Sonnet 3.7 的能力，在編碼和推理任務方面表現出色，具有改進的精度和可控性。在 SWE-bench 上達到 72.7% 的最先進性能，Sonnet 4 在能力和計算效率之間取得了平衡。"
     }
   },
-  "verified": false,
+  "verified": true,
   "websiteUrl": "https://www.anthropic.com",
   "docsUrl": "https://docs.anthropic.com/claude/docs/models-overview",
   "vendor": "Anthropic",
   "size": "Unknown",
   "contextWindow": 200000,
-  "maxOutput": 8000,
+  "maxOutput": 64000,
   "tokenPricing": {
     "input": 3,
-    "output": 3,
-    "cache": null
+    "output": 15,
+    "cache": 0.3
   },
-  "releaseDate": null,
-  "inputModalities": ["text"],
+  "releaseDate": "2025-05-22",
+  "lifecycle": "maintained",
+  "knowledgeCutoff": null,
+  "inputModalities": ["text", "image", "pdf"],
   "capabilities": ["function-calling", "tool-choice", "structured-outputs", "reasoning"],
   "benchmarks": {
     "sweBench": 64.93,

manifests/models/composer.json

@@ -51,6 +51,8 @@
     "cache": null
   },
   "releaseDate": null,
+  "lifecycle": "latest",
+  "knowledgeCutoff": null,
   "inputModalities": ["text"],
   "capabilities": ["function-calling", "tool-choice", "structured-outputs"],
   "benchmarks": {