@psnext/slingcli 2.5.20260710-1 → 2.5.20260718-2

package/node_modules/@earendil-works/pi-ai/dist/providers/zai-coding-cn.models.js

@@ -1,114 +1,5 @@
 // This file is auto-generated by scripts/generate-models.ts
 // Do not edit manually - run 'npm run generate-models' to update
-export const ZAI_CODING_CN_MODELS = {
-    "glm-4.5-air": {
-        id: "glm-4.5-air",
-        name: "GLM-4.5-Air",
-        api: "openai-completions",
-        provider: "zai-coding-cn",
-        baseUrl: "https://open.bigmodel.cn/api/coding/paas/v4",
-        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false, "thinkingFormat": "zai" },
-        reasoning: true,
-        input: ["text"],
-        cost: {
-            input: 0,
-            output: 0,
-            cacheRead: 0,
-            cacheWrite: 0,
-        },
-        contextWindow: 131072,
-        maxTokens: 98304,
-    },
-    "glm-4.7": {
-        id: "glm-4.7",
-        name: "GLM-4.7",
-        api: "openai-completions",
-        provider: "zai-coding-cn",
-        baseUrl: "https://open.bigmodel.cn/api/coding/paas/v4",
-        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false, "thinkingFormat": "zai", "zaiToolStream": true },
-        reasoning: true,
-        input: ["text"],
-        cost: {
-            input: 0,
-            output: 0,
-            cacheRead: 0,
-            cacheWrite: 0,
-        },
-        contextWindow: 204800,
-        maxTokens: 131072,
-    },
-    "glm-5-turbo": {
-        id: "glm-5-turbo",
-        name: "GLM-5-Turbo",
-        api: "openai-completions",
-        provider: "zai-coding-cn",
-        baseUrl: "https://open.bigmodel.cn/api/coding/paas/v4",
-        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false, "thinkingFormat": "zai", "zaiToolStream": true },
-        reasoning: true,
-        input: ["text"],
-        cost: {
-            input: 0,
-            output: 0,
-            cacheRead: 0,
-            cacheWrite: 0,
-        },
-        contextWindow: 200000,
-        maxTokens: 131072,
-    },
-    "glm-5.1": {
-        id: "glm-5.1",
-        name: "GLM-5.1",
-        api: "openai-completions",
-        provider: "zai-coding-cn",
-        baseUrl: "https://open.bigmodel.cn/api/coding/paas/v4",
-        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false, "thinkingFormat": "zai", "zaiToolStream": true },
-        reasoning: true,
-        input: ["text"],
-        cost: {
-            input: 0,
-            output: 0,
-            cacheRead: 0,
-            cacheWrite: 0,
-        },
-        contextWindow: 200000,
-        maxTokens: 131072,
-    },
-    "glm-5.2": {
-        id: "glm-5.2",
-        name: "GLM-5.2",
-        api: "openai-completions",
-        provider: "zai-coding-cn",
-        baseUrl: "https://open.bigmodel.cn/api/coding/paas/v4",
-        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": true, "thinkingFormat": "zai", "zaiToolStream": true },
-        reasoning: true,
-        thinkingLevelMap: { "minimal": null, "low": "high", "medium": "high", "high": "high", "xhigh": "max" },
-        input: ["text"],
-        cost: {
-            input: 0,
-            output: 0,
-            cacheRead: 0,
-            cacheWrite: 0,
-        },
-        contextWindow: 1000000,
-        maxTokens: 131072,
-    },
-    "glm-5v-turbo": {
-        id: "glm-5v-turbo",
-        name: "GLM-5V-Turbo",
-        api: "openai-completions",
-        provider: "zai-coding-cn",
-        baseUrl: "https://open.bigmodel.cn/api/coding/paas/v4",
-        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false, "thinkingFormat": "zai", "zaiToolStream": true },
-        reasoning: true,
-        input: ["text", "image"],
-        cost: {
-            input: 0,
-            output: 0,
-            cacheRead: 0,
-            cacheWrite: 0,
-        },
-        contextWindow: 200000,
-        maxTokens: 131072,
-    },
-};
+import values from "./data/zai-coding-cn.json" with { type: "json" };
+export const ZAI_CODING_CN_MODELS = values;
 //# sourceMappingURL=zai-coding-cn.models.js.map

package/node_modules/@earendil-works/pi-ai/dist/providers/zai.models.js

@@ -1,114 +1,5 @@
 // This file is auto-generated by scripts/generate-models.ts
 // Do not edit manually - run 'npm run generate-models' to update
-export const ZAI_MODELS = {
-    "glm-4.5-air": {
-        id: "glm-4.5-air",
-        name: "GLM-4.5-Air",
-        api: "openai-completions",
-        provider: "zai",
-        baseUrl: "https://api.z.ai/api/coding/paas/v4",
-        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false, "thinkingFormat": "zai" },
-        reasoning: true,
-        input: ["text"],
-        cost: {
-            input: 0,
-            output: 0,
-            cacheRead: 0,
-            cacheWrite: 0,
-        },
-        contextWindow: 131072,
-        maxTokens: 98304,
-    },
-    "glm-4.7": {
-        id: "glm-4.7",
-        name: "GLM-4.7",
-        api: "openai-completions",
-        provider: "zai",
-        baseUrl: "https://api.z.ai/api/coding/paas/v4",
-        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false, "thinkingFormat": "zai", "zaiToolStream": true },
-        reasoning: true,
-        input: ["text"],
-        cost: {
-            input: 0,
-            output: 0,
-            cacheRead: 0,
-            cacheWrite: 0,
-        },
-        contextWindow: 204800,
-        maxTokens: 131072,
-    },
-    "glm-5-turbo": {
-        id: "glm-5-turbo",
-        name: "GLM-5-Turbo",
-        api: "openai-completions",
-        provider: "zai",
-        baseUrl: "https://api.z.ai/api/coding/paas/v4",
-        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false, "thinkingFormat": "zai", "zaiToolStream": true },
-        reasoning: true,
-        input: ["text"],
-        cost: {
-            input: 0,
-            output: 0,
-            cacheRead: 0,
-            cacheWrite: 0,
-        },
-        contextWindow: 200000,
-        maxTokens: 131072,
-    },
-    "glm-5.1": {
-        id: "glm-5.1",
-        name: "GLM-5.1",
-        api: "openai-completions",
-        provider: "zai",
-        baseUrl: "https://api.z.ai/api/coding/paas/v4",
-        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false, "thinkingFormat": "zai", "zaiToolStream": true },
-        reasoning: true,
-        input: ["text"],
-        cost: {
-            input: 0,
-            output: 0,
-            cacheRead: 0,
-            cacheWrite: 0,
-        },
-        contextWindow: 200000,
-        maxTokens: 131072,
-    },
-    "glm-5.2": {
-        id: "glm-5.2",
-        name: "GLM-5.2",
-        api: "openai-completions",
-        provider: "zai",
-        baseUrl: "https://api.z.ai/api/coding/paas/v4",
-        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": true, "thinkingFormat": "zai", "zaiToolStream": true },
-        reasoning: true,
-        thinkingLevelMap: { "minimal": null, "low": "high", "medium": "high", "high": "high", "xhigh": "max" },
-        input: ["text"],
-        cost: {
-            input: 0,
-            output: 0,
-            cacheRead: 0,
-            cacheWrite: 0,
-        },
-        contextWindow: 1000000,
-        maxTokens: 131072,
-    },
-    "glm-5v-turbo": {
-        id: "glm-5v-turbo",
-        name: "GLM-5V-Turbo",
-        api: "openai-completions",
-        provider: "zai",
-        baseUrl: "https://api.z.ai/api/coding/paas/v4",
-        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false, "thinkingFormat": "zai", "zaiToolStream": true },
-        reasoning: true,
-        input: ["text", "image"],
-        cost: {
-            input: 0,
-            output: 0,
-            cacheRead: 0,
-            cacheWrite: 0,
-        },
-        contextWindow: 200000,
-        maxTokens: 131072,
-    },
-};
+import values from "./data/zai.json" with { type: "json" };
+export const ZAI_MODELS = values;
 //# sourceMappingURL=zai.models.js.map

package/node_modules/@earendil-works/pi-ai/dist/utils/deferred-tools.js

@@ -0,0 +1,36 @@
+const identityToolName = (name) => name;
+/** Split current tools into prefix and transcript-loaded definitions. */
+export function splitDeferredTools(context, enabled, normalizeName = identityToolName) {
+    const uniqueTools = new Map();
+    for (const tool of context.tools ?? [])
+        uniqueTools.set(normalizeName(tool.name), tool);
+    if (!enabled)
+        return { immediate: [...uniqueTools.values()], deferred: new Map() };
+    const deferredNames = new Set();
+    const usedNames = new Set();
+    for (const message of context.messages) {
+        if (message.role === "assistant") {
+            for (const block of message.content) {
+                if (block.type === "toolCall")
+                    usedNames.add(normalizeName(block.name));
+            }
+        }
+        else if (message.role === "toolResult") {
+            for (const name of message.addedToolNames ?? []) {
+                const normalizedName = normalizeName(name);
+                if (!usedNames.has(normalizedName))
+                    deferredNames.add(normalizedName);
+            }
+        }
+    }
+    const immediate = [];
+    const deferred = new Map();
+    for (const [name, tool] of uniqueTools) {
+        if (deferredNames.has(name))
+            deferred.set(name, tool);
+        else
+            immediate.push(tool);
+    }
+    return { immediate, deferred };
+}
+//# sourceMappingURL=deferred-tools.js.map

package/node_modules/@earendil-works/pi-ai/dist/utils/estimate.js

@@ -45,17 +45,25 @@ export function estimateMessageTokens(message) {
     return Math.ceil(chars / CHARS_PER_TOKEN);
 }
 function getLastAssistantUsageInfo(messages) {
-    for (let i = messages.length - 1; i >= 0; i--) {
+    let latestPrefixTimestamp = Number.NEGATIVE_INFINITY;
+    let usageInfo;
+    for (let i = 0; i < messages.length; i++) {
         const message = messages[i];
-        if (message.role !== "assistant")
-            continue;
-        const assistant = message;
-        if (assistant.stopReason === "aborted" || assistant.stopReason === "error")
-            continue;
-        if (calculateContextTokens(assistant.usage) > 0)
-            return { usage: assistant.usage, index: i };
+        if (message.role === "assistant") {
+            const assistant = message;
+            // A newer prefix message was inserted after this response (for example, a
+            // compaction summary), so its usage cannot describe the current prefix.
+            const usageAppliesToPrefix = assistant.timestamp >= latestPrefixTimestamp;
+            if (usageAppliesToPrefix &&
+                assistant.stopReason !== "aborted" &&
+                assistant.stopReason !== "error" &&
+                calculateContextTokens(assistant.usage) > 0) {
+                usageInfo = { usage: assistant.usage, index: i };
+            }
+        }
+        latestPrefixTimestamp = Math.max(latestPrefixTimestamp, message.timestamp);
     }
-    return undefined;
+    return usageInfo;
 }
 function estimateMessages(messages) {
     const usageInfo = getLastAssistantUsageInfo(messages);
@@ -72,6 +80,11 @@ function estimateMessages(messages) {
         tokens += estimateMessageTokens(message);
     return { tokens, usageTokens: 0, trailingTokens: tokens, lastUsageIndex: null };
 }
+function estimateToolsTokens(tools) {
+    if (!tools || tools.length === 0)
+        return 0;
+    return estimateTextTokens(safeJsonStringify(tools));
+}
 function isMessageArray(value) {
     return Array.isArray(value);
 }
@@ -79,12 +92,20 @@ export function estimateContextTokens(context) {
     if (isMessageArray(context))
         return estimateMessages(context);
     const estimate = estimateMessages(context.messages);
-    if (estimate.lastUsageIndex !== null)
-        return estimate;
-    let prefixTokens = context.systemPrompt ? estimateTextTokens(context.systemPrompt) : 0;
-    if (context.tools && context.tools.length > 0) {
-        prefixTokens += estimateTextTokens(safeJsonStringify(context.tools));
+    if (estimate.lastUsageIndex !== null) {
+        const addedNames = new Set(context.messages
+            .slice(estimate.lastUsageIndex + 1)
+            .filter((message) => message.role === "toolResult")
+            .flatMap((message) => message.addedToolNames ?? []));
+        const addedToolTokens = estimateToolsTokens(context.tools?.filter((tool) => addedNames.has(tool.name)));
+        return {
+            tokens: estimate.tokens + addedToolTokens,
+            usageTokens: estimate.usageTokens,
+            trailingTokens: estimate.trailingTokens + addedToolTokens,
+            lastUsageIndex: estimate.lastUsageIndex,
+        };
     }
+    const prefixTokens = (context.systemPrompt ? estimateTextTokens(context.systemPrompt) : 0) + estimateToolsTokens(context.tools);
     return {
         tokens: estimate.tokens + prefixTokens,
         usageTokens: estimate.usageTokens,

package/node_modules/@earendil-works/pi-ai/dist/utils/retry.js

@@ -46,6 +46,7 @@ const RETRYABLE_PROVIDER_ERROR_PATTERN = buildProviderErrorPattern([
     "upstream.?connect",
     "reset before headers",
     "socket hang up",
+    "socket connection was closed",
     "timed? out",
     "timeout",
     "terminated",
@@ -57,6 +58,7 @@ const RETRYABLE_PROVIDER_ERROR_PATTERN = buildProviderErrorPattern([
     // (#4433); Bedrock/Smithy can throw an HTTP/2 no-response error (#3594).
     "ended without",
     "stream ended before message_stop",
+    "stream ended before a terminal response event",
     "http2 request did not get a response",
     // Provider-requested retry delay cap failures should flow through the outer
     // retry policy so callers can surface/abort the backoff (#1123).
@@ -66,6 +68,8 @@ const RETRYABLE_PROVIDER_ERROR_PATTERN = buildProviderErrorPattern([
     "you can retry your request",
     "try your request again",
     "please retry your request",
+    // gRPC based providers (e.g. NVIDIA NIM)
+    "ResourceExhausted",
 ]);
 /**
  * Classifies whether a failed assistant message looks like a transient provider

package/node_modules/@earendil-works/pi-ai/dist/utils/validation.js

@@ -2,15 +2,6 @@ import { Compile } from "typebox/compile";
 import { Value } from "typebox/value";
 const validatorCache = new WeakMap();
 const TYPEBOX_KIND = Symbol.for("TypeBox.Kind");
-function isRecord(value) {
-    return typeof value === "object" && value !== null;
-}
-function isJsonSchemaObject(value) {
-    return isRecord(value);
-}
-function hasTypeBoxMetadata(schema) {
-    return isRecord(schema) && Object.getOwnPropertySymbols(schema).includes(TYPEBOX_KIND);
-}
 function getSchemaTypes(schema) {
     if (typeof schema.type === "string") {
         return [schema.type];
@@ -35,18 +26,12 @@ function matchesJsonType(value, type) {
         case "array":
             return Array.isArray(value);
         case "object":
-            return isRecord(value) && !Array.isArray(value);
+            return typeof value === "object" && value !== null && !Array.isArray(value);
         default:
             return false;
     }
 }
-function isValidatorSchema(value) {
-    return isRecord(value);
-}
 function getSubSchemaValidator(schema) {
-    if (!isValidatorSchema(schema)) {
-        return undefined;
-    }
     try {
         return getValidator(schema);
     }
@@ -138,7 +123,7 @@ function applySchemaObjectCoercion(value, schema) {
             value[key] = coerceWithJsonSchema(value[key], propertySchema);
         }
     }
-    if (schema.additionalProperties && isJsonSchemaObject(schema.additionalProperties)) {
+    if (schema.additionalProperties && typeof schema.additionalProperties === "object") {
         for (const [key, propertyValue] of Object.entries(value)) {
             if (definedKeys.has(key)) {
                 continue;
@@ -158,7 +143,7 @@ function applySchemaArrayCoercion(value, schema) {
         }
         return;
     }
-    if (isJsonSchemaObject(schema.items)) {
+    if (schema.items && typeof schema.items === "object") {
         for (let index = 0; index < value.length; index++) {
             value[index] = coerceWithJsonSchema(value[index], schema.items);
         }
@@ -199,7 +184,10 @@ function coerceWithJsonSchema(value, schema) {
             }
         }
     }
-    if (schemaTypes.includes("object") && isRecord(nextValue) && !Array.isArray(nextValue)) {
+    if (schemaTypes.includes("object") &&
+        typeof nextValue === "object" &&
+        nextValue !== null &&
+        !Array.isArray(nextValue)) {
         applySchemaObjectCoercion(nextValue, schema);
     }
     if (schemaTypes.includes("array") && Array.isArray(nextValue)) {
@@ -254,10 +242,10 @@ export function validateToolArguments(tool, toolCall) {
     const args = structuredClone(toolCall.arguments);
     Value.Convert(tool.parameters, args);
     const validator = getValidator(tool.parameters);
-    if (!hasTypeBoxMetadata(tool.parameters) && isJsonSchemaObject(tool.parameters)) {
+    if (!Object.getOwnPropertySymbols(tool.parameters).includes(TYPEBOX_KIND)) {
         const coerced = coerceWithJsonSchema(args, tool.parameters);
         if (coerced !== args) {
-            if (isRecord(args) && isRecord(coerced)) {
+            if (typeof args === "object" && args !== null && typeof coerced === "object" && coerced !== null) {
                 for (const key of Object.keys(args)) {
                     delete args[key];
                 }

package/node_modules/@earendil-works/pi-ai/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@earendil-works/pi-ai",
-	"version": "0.80.3",
+	"version": "0.80.10",
 	"description": "Unified LLM API with automatic model discovery and provider configuration",
 	"type": "module",
 	"main": "./dist/index.js",
@@ -34,6 +34,10 @@
 		"./bedrock-provider": {
 			"types": "./dist/bedrock-provider.d.ts",
 			"import": "./dist/bedrock-provider.js"
+		},
+		"./bun-oauth": {
+			"types": "./dist/bun-oauth.d.ts",
+			"import": "./dist/bun-oauth.js"
 		}
 	},
 	"bin": {
@@ -46,8 +50,9 @@
 	"scripts": {
 		"clean": "shx rm -rf dist",
 		"generate-models": "node scripts/generate-models.ts",
+		"generate-model-catalog": "node scripts/generate-models.ts --strict --json-only --json-output ../../.artifacts/model-catalog",
 		"generate-image-models": "node scripts/generate-image-models.ts",
-		"build": "npm run generate-models && npm run generate-image-models && tsgo -p tsconfig.build.json",
+		"build": "npm run generate-models && tsgo -p tsconfig.build.json && shx rm -rf dist/providers/data && shx cp -r src/providers/data dist/providers/data",
 		"test": "vitest --run",
 		"prepublishOnly": "npm run clean && npm run build"
 	},

package/node_modules/@earendil-works/pi-coding-agent/dist/bun/cli.js

@@ -1,7 +1,9 @@
 #!/usr/bin/env node
+import { registerBunOAuthFlows } from "@earendil-works/pi-ai/bun-oauth";
 import { APP_NAME } from "../config.js";
 process.title = APP_NAME;
 process.emitWarning = (() => { });
+registerBunOAuthFlows();
 import { restoreSandboxEnv } from "./restore-sandbox-env.js";
 restoreSandboxEnv();
 await import("./register-bedrock.js");

package/node_modules/@earendil-works/pi-coding-agent/dist/cli/args.js

@@ -3,7 +3,7 @@
  */
 import chalk from "chalk";
 import { APP_NAME, CONFIG_DIR_NAME, ENV_AGENT_DIR, ENV_SESSION_DIR } from "../config.js";
-const VALID_THINKING_LEVELS = ["off", "minimal", "low", "medium", "high", "xhigh"];
+const VALID_THINKING_LEVELS = ["off", "minimal", "low", "medium", "high", "xhigh", "max"];
 export function isValidThinkingLevel(level) {
     return VALID_THINKING_LEVELS.includes(level);
 }
@@ -217,10 +217,10 @@ ${chalk.bold("Commands:")}
   ${APP_NAME} install <source> [-l]     Install extension source and add to settings
   ${APP_NAME} remove <source> [-l]      Remove extension source from settings
   ${APP_NAME} uninstall <source> [-l]   Alias for remove
-  ${APP_NAME} update [source|self|pi]   Update pi (use --all for pi and extensions)
+  ${APP_NAME} update [source|self|pi]   Update pi, extensions, or model catalogs
   ${APP_NAME} list                      List installed extensions from settings
-  ${APP_NAME} config                    Open TUI to enable/disable package resources
-  ${APP_NAME} <command> --help          Show help for install/remove/uninstall/update/list
+  ${APP_NAME} config [-l]               Open TUI to enable/disable package resources (Tab switches scope)
+  ${APP_NAME} <command> --help          Show help for install/remove/uninstall/update/list/config
 
 ${chalk.bold("Options:")}
   --provider <name>              Provider name (default: google)
@@ -246,7 +246,7 @@ ${chalk.bold("Options:")}
                                  Applies to built-in, extension, and custom tools
   --exclude-tools, -xt <tools>   Comma-separated denylist of tool names to disable
                                  Applies to built-in, extension, and custom tools
-  --thinking <level>             Set thinking level: off, minimal, low, medium, high, xhigh
+  --thinking <level>             Set thinking level: off, minimal, low, medium, high, xhigh, max
   --extension, -e <path>         Load an extension file (can be used multiple times)
   --no-extensions, -ne           Disable extension discovery (explicit -e paths still work)
   --skill <path>                 Load a skill file or directory (can be used multiple times)

package/node_modules/@earendil-works/pi-coding-agent/dist/cli/config-selector.js

@@ -22,7 +22,7 @@ export async function selectConfig(options) {
             ui.stop();
             stopThemeWatcher();
             process.exit(0);
-        }, () => ui.requestRender(), ui.terminal.rows);
+        }, () => ui.requestRender(), ui.terminal.rows, options.writeScope, options.projectModeAvailable);
         ui.addChild(selector);
         ui.setFocus(selector.getResourceList());
         ui.start();

package/node_modules/@earendil-works/pi-coding-agent/dist/cli/list-models.js

@@ -21,12 +21,12 @@ function formatTokenCount(count) {
 /**
  * List available models, optionally filtered by search pattern
  */
-export async function listModels(modelRegistry, searchPattern) {
-    const loadError = modelRegistry.getError();
+export async function listModels(modelRuntime, searchPattern) {
+    const loadError = modelRuntime.getError();
     if (loadError) {
         console.error(chalk.yellow(`Warning: errors loading models.json:\n${loadError}`));
     }
-    const models = modelRegistry.getAvailable();
+    const models = [...(await modelRuntime.getAvailable())];
     if (models.length === 0) {
         console.log(formatNoModelsAvailableMessage());
         return;

package/node_modules/@earendil-works/pi-coding-agent/dist/core/agent-session-runtime.js

@@ -210,6 +210,9 @@ export class AgentSessionRuntime {
                 await this.finishSessionReplacement(options?.withSession);
                 return { cancelled: false, selectedText };
             }
+            if (!existsSync(currentSessionFile)) {
+                throw new Error("This session has not been saved yet. Wait for the first assistant response before cloning or forking it.");
+            }
             const sessionManager = SessionManager.open(currentSessionFile, sessionDir);
             const forkedSessionPath = sessionManager.createBranchedSession(targetLeafId);
             if (!forkedSessionPath) {

package/node_modules/@earendil-works/pi-coding-agent/dist/core/agent-session-services.js

@@ -1,8 +1,7 @@
 import { join } from "node:path";
 import { getAgentDir } from "../config.js";
 import { resolvePath } from "../utils/paths.js";
-import { AuthStorage } from "./auth-storage.js";
-import { ModelRegistry } from "./model-registry.js";
+import { ModelRuntime } from "./model-runtime.js";
 import { DefaultResourceLoader, } from "./resource-loader.js";
 import { createAgentSession } from "./sdk.js";
 import { SettingsManager } from "./settings-manager.js";
@@ -54,9 +53,12 @@ function applyExtensionFlagValues(resourceLoader, extensionFlagValues) {
 export async function createAgentSessionServices(options) {
     const cwd = resolvePath(options.cwd);
     const agentDir = options.agentDir ? resolvePath(options.agentDir) : getAgentDir();
-    const authStorage = options.authStorage ?? AuthStorage.create(join(agentDir, "auth.json"));
+    const modelRuntime = options.modelRuntime ??
+        (await ModelRuntime.create({
+            authPath: join(agentDir, "auth.json"),
+            modelsPath: join(agentDir, "models.json"),
+        }));
     const settingsManager = options.settingsManager ?? SettingsManager.create(cwd, agentDir);
-    const modelRegistry = options.modelRegistry ?? ModelRegistry.create(authStorage, join(agentDir, "models.json"));
     const resourceLoader = new DefaultResourceLoader({
         ...(options.resourceLoaderOptions ?? {}),
         cwd,
@@ -68,7 +70,7 @@ export async function createAgentSessionServices(options) {
     const extensionsResult = resourceLoader.getExtensions();
     for (const { name, config, extensionPath } of extensionsResult.runtime.pendingProviderRegistrations) {
         try {
-            modelRegistry.registerProvider(name, config);
+            modelRuntime.registerProvider(name, config);
         }
         catch (error) {
             const message = error instanceof Error ? error.message : String(error);
@@ -79,13 +81,26 @@ export async function createAgentSessionServices(options) {
         }
     }
     extensionsResult.runtime.pendingProviderRegistrations = [];
+    for (const { provider, extensionPath } of extensionsResult.runtime.pendingNativeProviderRegistrations) {
+        try {
+            modelRuntime.registerNativeProvider(provider);
+        }
+        catch (error) {
+            const message = error instanceof Error ? error.message : String(error);
+            diagnostics.push({
+                type: "error",
+                message: `Extension "${extensionPath}" error: ${message}`,
+            });
+        }
+    }
+    extensionsResult.runtime.pendingNativeProviderRegistrations = [];
+    await modelRuntime.refresh({ allowNetwork: false });
     diagnostics.push(...applyExtensionFlagValues(resourceLoader, options.extensionFlagValues));
     return {
         cwd,
         agentDir,
-        authStorage,
+        modelRuntime,
         settingsManager,
-        modelRegistry,
         resourceLoader,
         diagnostics,
     };
@@ -101,9 +116,8 @@ export async function createAgentSessionFromServices(options) {
     return createAgentSession({
         cwd: options.services.cwd,
         agentDir: options.services.agentDir,
-        authStorage: options.services.authStorage,
+        modelRuntime: options.services.modelRuntime,
         settingsManager: options.services.settingsManager,
-        modelRegistry: options.services.modelRegistry,
         resourceLoader: options.services.resourceLoader,
         sessionManager: options.sessionManager,
         model: options.model,