@x1a0f3n9/dsh-llm 0.1.5-rc.2 → 0.1.5-rc.4

package/lib/index.js

@@ -155,6 +155,12 @@ const STRUCTURED_CONTEXT_OVERFLOW = new RegExp(String.raw`(?:^|[^a-z0-9])context
 const TOO_LARGE_FOR_CONTEXT = new RegExp(String.raw`\b(?:request|prompt|input|messages?)\s+(?:is\s+|are\s+)?` + String.raw`too\s+(?:large|long)\s+for\s+(?:(?:this|the)\s+)?` + String.raw`(?:model(?:'s)?\s+)?context(?:\s+window)?\b`, "i");
 /** "Exceeds" wording is safe only when its object is explicitly the model context. */
 const EXCEEDS_MODEL_CONTEXT = new RegExp(String.raw`\b(?:input|prompt|request|messages?)\b.{0,40}` + String.raw`\b(?:exceed(?:s|ed)?|overflows?|is\s+larger\s+than)\b.{0,40}` + String.raw`\b(?:the\s+)?(?:model(?:'s)?\s+)?context(?:\s+(?:length|window))?\b`, "i");
+/** xAI Grok and similar OpenAI-compatible gateways name the bound as prompt length. */
+const MAXIMUM_PROMPT_LENGTH = /\bmaximum prompt length\b|\bprompt is too long\b/i;
+/** Groq and similar gateways ask the caller to shorten the message list. */
+const REDUCE_MESSAGE_LENGTH = /\breduce the length of the messages\b/i;
+/** Chinese OpenAI-compatible gateways often return localized overflow copy. */
+const CHINESE_CONTEXT_OVERFLOW = /上下文(?:窗口|长度)|(?:提示词|输入|请求).{0,8}过长/;
 /**
 * Recognize the context-overflow wording used by OpenAI-compatible providers
 * and library adapters. Adapters pass all available provider code, type, and
@@ -163,7 +169,7 @@ const EXCEEDS_MODEL_CONTEXT = new RegExp(String.raw`\b(?:input|prompt|request|me
 * @returns true when the detail identifies a request exceeding the model context window.
 */
 function isContextWindowExceededError(detail) {
-	return STRUCTURED_CONTEXT_OVERFLOW.test(detail) || /\b(?:maximum|max)(?:\s+(?:allowed|supported))?\s+context\s+(?:length|window)\b/i.test(detail) || TOO_LARGE_FOR_CONTEXT.test(detail) || /\b(?:input|prompt|request)\s+(?:is\s+)?too\s+(?:long|large)\s+for\s+(?:this|the)\s+model\b/i.test(detail) || EXCEEDS_MODEL_CONTEXT.test(detail);
+	return STRUCTURED_CONTEXT_OVERFLOW.test(detail) || /\b(?:maximum|max)(?:\s+(?:allowed|supported))?\s+context\s+(?:length|window)\b/i.test(detail) || TOO_LARGE_FOR_CONTEXT.test(detail) || /\b(?:input|prompt|request)\s+(?:is\s+)?too\s+(?:long|large)\s+for\s+(?:this|the)\s+model\b/i.test(detail) || /\btoo long for (?:this|the) model\b/i.test(detail) || EXCEEDS_MODEL_CONTEXT.test(detail) || MAXIMUM_PROMPT_LENGTH.test(detail) || REDUCE_MESSAGE_LENGTH.test(detail) || CHINESE_CONTEXT_OVERFLOW.test(detail);
 }
 /**
 * Recognize provider wording that identifies an exhausted account quota rather
@@ -172,7 +178,7 @@ function isContextWindowExceededError(detail) {
 * @returns true only for terminal quota, balance, credit, budget, or usage-limit wording.
 */
 function isQuotaExceededError(detail) {
-	return /\binsufficient[\s_-]+(?:quota|balance|credits?)\b/i.test(detail) || /\b(?:quota|usage[\s_-]+limit)[\s_-]+(?:exceeded|exhausted|reached)\b/i.test(detail) || /\bexceed(?:ed|s)?[\s_-]+(?:(?:your|the)[\s_-]+)?(?:current[\s_-]+)?quota\b/i.test(detail) || /\b(?:balance|credits?)[\s_-]+(?:exhausted|depleted)\b/i.test(detail) || /\bout[\s_-]+of[\s_-]+(?:credits?|budget)\b/i.test(detail);
+	return /\binsufficient[\s_-]+(?:quota|balance|credits?)\b/i.test(detail) || /\b(?:quota|usage[\s_-]+limit)[\s_-]+(?:exceeded|exhausted|reached)\b/i.test(detail) || /\bexceed(?:ed|s)?[\s_-]+(?:(?:your|the)[\s_-]+)?(?:current[\s_-]+)?quota\b/i.test(detail) || /\b(?:balance|credits?)[\s_-]+(?:exhausted|depleted)\b/i.test(detail) || /\bout[\s_-]+of[\s_-]+(?:credits?|budget)\b/i.test(detail) || /\bexceeded the .{0,24}usage[\s_-]+quota\b/i.test(detail) || /\breached your usage limit\b/i.test(detail) || /\busage[\s_-]+limit\b.{0,48}\bbilling cycle\b/i.test(detail);
 }
 /**
 * Render a thrown value with its full `cause` chain and AggregateError
@@ -1953,7 +1959,8 @@ let LlmRuntime = (() => {
 					id: model.id,
 					...model.name === void 0 ? {} : { name: model.name },
 					...model.contextWindow === void 0 ? {} : { contextWindow: model.contextWindow },
-					...model.maxTokens === void 0 ? {} : { maxTokens: model.maxTokens }
+					...model.maxTokens === void 0 ? {} : { maxTokens: model.maxTokens },
+					...model.inputModalities === void 0 ? {} : { inputModalities: model.inputModalities }
 				});
 			}
 			return models;

package/lib/typert.host.js

@@ -13,6 +13,7 @@ const _x1a0f3n9_dsh_llm_llm_discoverModels_result$schema = z.array(z.object({
   'name': z.string().optional(),
   'contextWindow': z.number().optional(),
   'maxTokens': z.number().optional(),
+  'inputModalities': z.array(z.union([z.literal("text"), z.literal("image")])).optional(),
 }))
 const _x1a0f3n9_dsh_llm_llm_listConfigurableProviders_result$schema = z.array(z.object({
   'provider': z.string(),
@@ -68,7 +69,7 @@ export const TYPERT = {
         typeSymbol: '@x1a0f3n9/dsh-llm#llm/discoverModels:result',
         schema: _x1a0f3n9_dsh_llm_llm_discoverModels_result$schema,
       },
-      sourceLocation: {"file":"packages/llm/llm/src/index.ts","line":628,"column":9},
+      sourceLocation: {"file":"packages/llm/llm/src/index.ts","line":629,"column":9},
     },
     {
       id: '@x1a0f3n9/dsh-llm#llm/listConfigurableProviders',
@@ -324,7 +325,7 @@ export const TYPERT = {
           },
           {
             "name": "LlmDiscoveredModel",
-            "declaration": "export interface LlmDiscoveredModel {\n    id: string;\n    name?: string;\n    contextWindow?: number;\n    maxTokens?: number;\n}"
+            "declaration": "export interface LlmDiscoveredModel {\n    id: string;\n    name?: string;\n    contextWindow?: number;\n    maxTokens?: number;\n    inputModalities?: readonly ModelModality[];\n}"
           },
           {
             "name": "LlmFailure",

package/lib/typert.remote-client.js

@@ -13,6 +13,7 @@ const _x1a0f3n9_dsh_llm_llm_discoverModels_result$schema = z.array(z.object({
   'name': z.string().optional(),
   'contextWindow': z.number().optional(),
   'maxTokens': z.number().optional(),
+  'inputModalities': z.array(z.union([z.literal("text"), z.literal("image")])).optional(),
 }))
 const _x1a0f3n9_dsh_llm_llm_listConfigurableProviders_result$schema = z.array(z.object({
   'provider': z.string(),
@@ -65,7 +66,7 @@ export const TYPERT_REMOTE = {
         typeSymbol: '@x1a0f3n9/dsh-llm#llm/discoverModels:result',
         schema: _x1a0f3n9_dsh_llm_llm_discoverModels_result$schema,
       },
-      sourceLocation: {"file":"packages/llm/llm/src/index.ts","line":628,"column":9},
+      sourceLocation: {"file":"packages/llm/llm/src/index.ts","line":629,"column":9},
     },
     {
       id: '@x1a0f3n9/dsh-llm#llm/listConfigurableProviders',

package/lib/types/error.js

@@ -51,6 +51,12 @@ const TOO_LARGE_FOR_CONTEXT = new RegExp(String.raw `\b(?:request|prompt|input|m
 const EXCEEDS_MODEL_CONTEXT = new RegExp(String.raw `\b(?:input|prompt|request|messages?)\b.{0,40}`
     + String.raw `\b(?:exceed(?:s|ed)?|overflows?|is\s+larger\s+than)\b.{0,40}`
     + String.raw `\b(?:the\s+)?(?:model(?:'s)?\s+)?context(?:\s+(?:length|window))?\b`, 'i');
+/** xAI Grok and similar OpenAI-compatible gateways name the bound as prompt length. */
+const MAXIMUM_PROMPT_LENGTH = /\bmaximum prompt length\b|\bprompt is too long\b/i;
+/** Groq and similar gateways ask the caller to shorten the message list. */
+const REDUCE_MESSAGE_LENGTH = /\breduce the length of the messages\b/i;
+/** Chinese OpenAI-compatible gateways often return localized overflow copy. */
+const CHINESE_CONTEXT_OVERFLOW = /上下文(?:窗口|长度)|(?:提示词|输入|请求).{0,8}过长/;
 /**
  * Recognize the context-overflow wording used by OpenAI-compatible providers
  * and library adapters. Adapters pass all available provider code, type, and
@@ -63,7 +69,11 @@ export function isContextWindowExceededError(detail) {
         || /\b(?:maximum|max)(?:\s+(?:allowed|supported))?\s+context\s+(?:length|window)\b/i.test(detail)
         || TOO_LARGE_FOR_CONTEXT.test(detail)
         || /\b(?:input|prompt|request)\s+(?:is\s+)?too\s+(?:long|large)\s+for\s+(?:this|the)\s+model\b/i.test(detail)
-        || EXCEEDS_MODEL_CONTEXT.test(detail);
+        || /\btoo long for (?:this|the) model\b/i.test(detail)
+        || EXCEEDS_MODEL_CONTEXT.test(detail)
+        || MAXIMUM_PROMPT_LENGTH.test(detail)
+        || REDUCE_MESSAGE_LENGTH.test(detail)
+        || CHINESE_CONTEXT_OVERFLOW.test(detail);
 }
 /**
  * Recognize provider wording that identifies an exhausted account quota rather
@@ -76,7 +86,10 @@ export function isQuotaExceededError(detail) {
         || /\b(?:quota|usage[\s_-]+limit)[\s_-]+(?:exceeded|exhausted|reached)\b/i.test(detail)
         || /\bexceed(?:ed|s)?[\s_-]+(?:(?:your|the)[\s_-]+)?(?:current[\s_-]+)?quota\b/i.test(detail)
         || /\b(?:balance|credits?)[\s_-]+(?:exhausted|depleted)\b/i.test(detail)
-        || /\bout[\s_-]+of[\s_-]+(?:credits?|budget)\b/i.test(detail);
+        || /\bout[\s_-]+of[\s_-]+(?:credits?|budget)\b/i.test(detail)
+        || /\bexceeded the .{0,24}usage[\s_-]+quota\b/i.test(detail)
+        || /\breached your usage limit\b/i.test(detail)
+        || /\busage[\s_-]+limit\b.{0,48}\bbilling cycle\b/i.test(detail);
 }
 /**
  * Render a thrown value with its full `cause` chain and AggregateError

package/lib/types/index.js

@@ -487,6 +487,7 @@ let LlmRuntime = (() => {
                     ...model.name === undefined ? {} : { name: model.name },
                     ...model.contextWindow === undefined ? {} : { contextWindow: model.contextWindow },
                     ...model.maxTokens === undefined ? {} : { maxTokens: model.maxTokens },
+                    ...model.inputModalities === undefined ? {} : { inputModalities: model.inputModalities },
                 });
             }
             return models;

package/lib/types/types.d.ts

@@ -272,6 +272,11 @@ export interface LlmDiscoveredModel {
     contextWindow?: number;
     /** Maximum output tokens, when disclosed. */
     maxTokens?: number;
+    /**
+     * Accepted request modalities when the listing or an installed catalog
+     * discloses them. Absent means unknown, not text-only.
+     */
+    inputModalities?: readonly ModelModality[];
 }
 /** One adapter-discovered model; catalog membership is advisory, not request validation. */
 export interface LlmModelInfo {

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "@x1a0f3n9/dsh-llm",
   "description": "Provider-neutral LLM service interface for the DeepSeek Harness",
-  "version": "0.1.5-rc.2",
+  "version": "0.1.5-rc.4",
   "publishConfig": {
     "access": "public"
   },
@@ -65,16 +65,16 @@
   },
   "dependencies": {
     "zod": "^4.4.3",
-    "@x1a0f3n9/dsh-util-crypto": "^0.1.5-rc.2",
+    "@x1a0f3n9/dsh-util-crypto": "^0.1.5-rc.4",
     "@deepseek-ai/schemastery": "^3.18.2",
-    "@x1a0f3n9/dsh-timeout": "^0.1.5-rc.2",
-    "@x1a0f3n9/dsh-brand": "^0.1.5-rc.2",
-    "@x1a0f3n9/dsh-util-values": "^0.1.5-rc.2",
-    "@x1a0f3n9/dsh-typert-protocol": "^0.1.5-rc.2"
+    "@x1a0f3n9/dsh-timeout": "^0.1.5-rc.4",
+    "@x1a0f3n9/dsh-brand": "^0.1.5-rc.4",
+    "@x1a0f3n9/dsh-util-values": "^0.1.5-rc.4",
+    "@x1a0f3n9/dsh-typert-protocol": "^0.1.5-rc.4"
   },
   "devDependencies": {
-    "@x1a0f3n9/dsh-attachment": "^0.1.5-rc.2",
-    "@x1a0f3n9/dsh-invariants": "^0.1.5-rc.2",
+    "@x1a0f3n9/dsh-attachment": "^0.1.5-rc.4",
+    "@x1a0f3n9/dsh-invariants": "^0.1.5-rc.4",
     "@deepseek-ai/cordis": "^4.0.2"
   }
 }