fix: Rename Task to LLMTask and improve api key required message (#652)

* fix: Rename task to llmtask and improve api key required message

* cr

* cr

apps/open-swe/src/graphs/manager/nodes/classify-message/index.ts

@@ -14,8 +14,8 @@ import { z } from "zod";
 import {
   loadModel,
   supportsParallelToolCallsParam,
-  Task,
 } from "../../../../utils/llms/index.js";
+import { LLMTask } from "@open-swe/shared/open-swe/llm-task";
 import { Command, END } from "@langchain/langgraph";
 import { getMessageContentString } from "@open-swe/shared/messages";
 import {
@@ -108,10 +108,10 @@ export async function classifyMessage(
     description: "Respond to the user's message and determine how to route it.",
     schema,
   };
-  const model = await loadModel(config, Task.ROUTER);
+  const model = await loadModel(config, LLMTask.ROUTER);
   const modelSupportsParallelToolCallsParam = supportsParallelToolCallsParam(
     config,
-    Task.ROUTER,
+    LLMTask.ROUTER,
   );
   const modelWithTools = model.bindTools([respondAndRouteTool], {
     tool_choice: respondAndRouteTool.name,

apps/open-swe/src/graphs/manager/utils/generate-issue-fields.ts

@@ -4,15 +4,15 @@ import { z } from "zod";
 import {
   loadModel,
   supportsParallelToolCallsParam,
-  Task,
 } from "../../../utils/llms/index.js";
+import { LLMTask } from "@open-swe/shared/open-swe/llm-task";
 import { getMessageString } from "../../../utils/message/content.js";
 
 export async function createIssueFieldsFromMessages(
   messages: BaseMessage[],
   configurable: GraphConfig["configurable"],
 ): Promise<{ title: string; body: string }> {
-  const model = await loadModel({ configurable }, Task.ROUTER);
+  const model = await loadModel({ configurable }, LLMTask.ROUTER);
   const githubIssueTool = {
     name: "create_github_issue",
     description: "Create a new GitHub issue with the given title and body.",
@@ -31,7 +31,7 @@ export async function createIssueFieldsFromMessages(
   };
   const modelSupportsParallelToolCallsParam = supportsParallelToolCallsParam(
     { configurable },
-    Task.ROUTER,
+    LLMTask.ROUTER,
   );
   const modelWithTools = model
     .bindTools([githubIssueTool], {

apps/open-swe/src/graphs/planner/nodes/determine-needs-context.ts

@@ -8,8 +8,8 @@ import { z } from "zod";
 import {
   loadModel,
   supportsParallelToolCallsParam,
-  Task,
 } from "../../../utils/llms/index.js";
+import { LLMTask } from "@open-swe/shared/open-swe/llm-task";
 import { getMissingMessages } from "../../../utils/github/issue-messages.js";
 import { getMessageString } from "../../../utils/message/content.js";
 import { isHumanMessage } from "@langchain/core/messages";
@@ -119,10 +119,10 @@ export async function determineNeedsContext(
 ): Promise<Command> {
   const [missingMessages, model] = await Promise.all([
     getMissingMessages(state, config),
-    loadModel(config, Task.ROUTER),
+    loadModel(config, LLMTask.ROUTER),
   ]);
   const modelManager = getModelManager();
-  const modelName = modelManager.getModelNameForTask(config, Task.ROUTER);
+  const modelName = modelManager.getModelNameForTask(config, LLMTask.ROUTER);
   if (!missingMessages.length) {
     throw new Error(
       "Can not determine if more context is needed if there are no missing messages.",
@@ -130,7 +130,7 @@ export async function determineNeedsContext(
   }
   const modelSupportsParallelToolCallsParam = supportsParallelToolCallsParam(
     config,
-    Task.ROUTER,
+    LLMTask.ROUTER,
   );
   const modelWithTools = model.bindTools([determineContextTool], {
     tool_choice: determineContextTool.name,

apps/open-swe/src/graphs/planner/nodes/generate-message/index.ts

@@ -2,8 +2,8 @@ import {
   getModelManager,
   loadModel,
   supportsParallelToolCallsParam,
-  Task,
 } from "../../../../utils/llms/index.js";
+import { LLMTask } from "@open-swe/shared/open-swe/llm-task";
 import {
   createGetURLContentTool,
   createShellTool,
@@ -86,12 +86,12 @@ export async function generateAction(
   state: PlannerGraphState,
   config: GraphConfig,
 ): Promise<PlannerGraphUpdate> {
-  const model = await loadModel(config, Task.PLANNER);
+  const model = await loadModel(config, LLMTask.PLANNER);
   const modelManager = getModelManager();
-  const modelName = modelManager.getModelNameForTask(config, Task.PLANNER);
+  const modelName = modelManager.getModelNameForTask(config, LLMTask.PLANNER);
   const modelSupportsParallelToolCallsParam = supportsParallelToolCallsParam(
     config,
-    Task.PLANNER,
+    LLMTask.PLANNER,
   );
   const mcpTools = await getMcpTools(config);
 

apps/open-swe/src/graphs/planner/nodes/generate-plan/index.ts

@@ -5,8 +5,8 @@ import { GraphConfig } from "@open-swe/shared/open-swe/types";
 import {
   loadModel,
   supportsParallelToolCallsParam,
-  Task,
 } from "../../../../utils/llms/index.js";
+import { LLMTask } from "@open-swe/shared/open-swe/llm-task";
 import {
   PlannerGraphState,
   PlannerGraphUpdate,
@@ -55,12 +55,12 @@ export async function generatePlan(
   state: PlannerGraphState,
   config: GraphConfig,
 ): Promise<PlannerGraphUpdate> {
-  const model = await loadModel(config, Task.PLANNER);
+  const model = await loadModel(config, LLMTask.PLANNER);
   const modelManager = getModelManager();
-  const modelName = modelManager.getModelNameForTask(config, Task.PLANNER);
+  const modelName = modelManager.getModelNameForTask(config, LLMTask.PLANNER);
   const modelSupportsParallelToolCallsParam = supportsParallelToolCallsParam(
     config,
-    Task.PLANNER,
+    LLMTask.PLANNER,
   );
   const sessionPlanTool = createSessionPlanToolFields();
   const modelWithTools = model.bindTools([sessionPlanTool], {

apps/open-swe/src/graphs/planner/nodes/notetaker.ts

@@ -8,8 +8,8 @@ import {
 import {
   loadModel,
   supportsParallelToolCallsParam,
-  Task,
 } from "../../../utils/llms/index.js";
+import { LLMTask } from "@open-swe/shared/open-swe/llm-task";
 import { getMessageString } from "../../../utils/message/content.js";
 import { formatUserRequestPrompt } from "../../../utils/user-request.js";
 import { formatCustomRulesPrompt } from "../../../utils/custom-rules.js";
@@ -102,12 +102,15 @@ export async function notetaker(
   state: PlannerGraphState,
   config: GraphConfig,
 ): Promise<PlannerGraphUpdate> {
-  const model = await loadModel(config, Task.SUMMARIZER);
+  const model = await loadModel(config, LLMTask.SUMMARIZER);
   const modelManager = getModelManager();
-  const modelName = modelManager.getModelNameForTask(config, Task.SUMMARIZER);
+  const modelName = modelManager.getModelNameForTask(
+    config,
+    LLMTask.SUMMARIZER,
+  );
   const modelSupportsParallelToolCallsParam = supportsParallelToolCallsParam(
     config,
-    Task.SUMMARIZER,
+    LLMTask.SUMMARIZER,
   );
   const modelWithTools = model.bindTools([condenseContextTool], {
     tool_choice: condenseContextTool.name,

apps/open-swe/src/graphs/planner/nodes/rewrite-plan.ts

@@ -16,8 +16,8 @@ import {
 import {
   loadModel,
   supportsParallelToolCallsParam,
-  Task,
 } from "../../../utils/llms/index.js";
+import { LLMTask } from "@open-swe/shared/open-swe/llm-task";
 import { FallbackRunnable } from "../../../utils/runtime-fallback.js";
 
 const systemPromptIdentifyChanges = `You are operating as an agentic coding assistant built by LangChain. You've previously been given a task to generate a plan of action for, to address the user's initial request.
@@ -265,10 +265,10 @@ export async function rewritePlan(
     throw new Error("No plan change request found.");
   }
 
-  const model = await loadModel(config, Task.PLANNER);
+  const model = await loadModel(config, LLMTask.PLANNER);
   const modelSupportsParallelToolCallsParam = supportsParallelToolCallsParam(
     config,
-    Task.PLANNER,
+    LLMTask.PLANNER,
   );
   const tasksToModify = await identifyTasksToModify(
     state,

apps/open-swe/src/graphs/programmer/nodes/diagnose-error.ts

@@ -17,8 +17,8 @@ import { getMessageContentString } from "@open-swe/shared/messages";
 import {
   loadModel,
   supportsParallelToolCallsParam,
-  Task,
 } from "../../../utils/llms/index.js";
+import { LLMTask } from "@open-swe/shared/open-swe/llm-task";
 import { z } from "zod";
 import { createLogger, LogLevel } from "../../../utils/logger.js";
 import {
@@ -108,10 +108,10 @@ export async function diagnoseError(
 
   logger.info("The last two tool calls resulted in errors. Diagnosing error.");
 
-  const model = await loadModel(config, Task.SUMMARIZER);
+  const model = await loadModel(config, LLMTask.SUMMARIZER);
   const modelSupportsParallelToolCallsParam = supportsParallelToolCallsParam(
     config,
-    Task.SUMMARIZER,
+    LLMTask.SUMMARIZER,
   );
   const modelWithTools = model.bindTools([diagnoseErrorTool], {
     tool_choice: diagnoseErrorTool.name,

apps/open-swe/src/graphs/programmer/nodes/generate-conclusion.ts

@@ -4,7 +4,8 @@ import {
   GraphUpdate,
   PlanItem,
 } from "@open-swe/shared/open-swe/types";
-import { loadModel, Task } from "../../../utils/llms/index.js";
+import { loadModel } from "../../../utils/llms/index.js";
+import { LLMTask } from "@open-swe/shared/open-swe/llm-task";
 import { getMessageContentString } from "@open-swe/shared/messages";
 import { getMessageString } from "../../../utils/message/content.js";
 import { createLogger, LogLevel } from "../../../utils/logger.js";
@@ -40,9 +41,12 @@ export async function generateConclusion(
   state: GraphState,
   config: GraphConfig,
 ): Promise<GraphUpdate> {
-  const model = await loadModel(config, Task.SUMMARIZER);
+  const model = await loadModel(config, LLMTask.SUMMARIZER);
   const modelManager = getModelManager();
-  const modelName = modelManager.getModelNameForTask(config, Task.SUMMARIZER);
+  const modelName = modelManager.getModelNameForTask(
+    config,
+    LLMTask.SUMMARIZER,
+  );
 
   const userRequestPrompt = formatUserRequestPrompt(state.messages);
   const userMessage = `${userRequestPrompt}

apps/open-swe/src/graphs/programmer/nodes/generate-message/index.ts

@@ -10,8 +10,8 @@ import {
   loadModel,
   Provider,
   supportsParallelToolCallsParam,
-  Task,
 } from "../../../../utils/llms/index.js";
+import { LLMTask } from "@open-swe/shared/open-swe/llm-task";
 import {
   createShellTool,
   createApplyPatchTool,
@@ -264,10 +264,13 @@ export async function generateAction(
   config: GraphConfig,
 ): Promise<GraphUpdate> {
   const modelManager = getModelManager();
-  const modelName = modelManager.getModelNameForTask(config, Task.PROGRAMMER);
+  const modelName = modelManager.getModelNameForTask(
+    config,
+    LLMTask.PROGRAMMER,
+  );
   const modelSupportsParallelToolCallsParam = supportsParallelToolCallsParam(
     config,
-    Task.PROGRAMMER,
+    LLMTask.PROGRAMMER,
   );
   const markTaskCompletedTool = createMarkTaskCompletedToolFields();
   const isAnthropicModel = modelName.includes("claude-");
@@ -286,7 +289,7 @@ export async function generateAction(
     },
   );
 
-  const model = await loadModel(config, Task.PROGRAMMER, {
+  const model = await loadModel(config, LLMTask.PROGRAMMER, {
     providerTools: providerTools,
     providerMessages: providerMessages,
   });

apps/open-swe/src/graphs/programmer/nodes/open-pr.ts

@@ -21,8 +21,8 @@ import { z } from "zod";
 import {
   loadModel,
   supportsParallelToolCallsParam,
-  Task,
 } from "../../../utils/llms/index.js";
+import { LLMTask } from "@open-swe/shared/open-swe/llm-task";
 import { formatPlanPromptWithSummaries } from "../../../utils/plan-prompt.js";
 import { formatUserRequestPrompt } from "../../../utils/user-request.js";
 import { AIMessage, BaseMessage, ToolMessage } from "@langchain/core/messages";
@@ -138,12 +138,12 @@ export async function openPullRequest(
 
   const openPrTool = createOpenPrToolFields();
   // use the router model since this is a simple task that doesn't need an advanced model
-  const model = await loadModel(config, Task.ROUTER);
+  const model = await loadModel(config, LLMTask.ROUTER);
   const modelManager = getModelManager();
-  const modelName = modelManager.getModelNameForTask(config, Task.ROUTER);
+  const modelName = modelManager.getModelNameForTask(config, LLMTask.ROUTER);
   const modelSupportsParallelToolCallsParam = supportsParallelToolCallsParam(
     config,
-    Task.ROUTER,
+    LLMTask.ROUTER,
   );
   const modelWithTool = model.bindTools([openPrTool], {
     tool_choice: openPrTool.name,

apps/open-swe/src/graphs/programmer/nodes/summarize-history.ts

@@ -5,7 +5,8 @@ import {
   GraphUpdate,
   PlanItem,
 } from "@open-swe/shared/open-swe/types";
-import { loadModel, Task } from "../../../utils/llms/index.js";
+import { loadModel } from "../../../utils/llms/index.js";
+import { LLMTask } from "@open-swe/shared/open-swe/llm-task";
 import {
   AIMessage,
   BaseMessage,
@@ -148,9 +149,12 @@ export async function summarizeHistory(
   state: GraphState,
   config: GraphConfig,
 ): Promise<GraphUpdate> {
-  const model = await loadModel(config, Task.SUMMARIZER);
+  const model = await loadModel(config, LLMTask.SUMMARIZER);
   const modelManager = getModelManager();
-  const modelName = modelManager.getModelNameForTask(config, Task.SUMMARIZER);
+  const modelName = modelManager.getModelNameForTask(
+    config,
+    LLMTask.SUMMARIZER,
+  );
 
   const plan = getActivePlanItems(state.taskPlan);
   const conversationHistoryToSummarize = await getMessagesSinceLastSummary(

apps/open-swe/src/graphs/programmer/nodes/update-plan.ts

@@ -9,8 +9,8 @@ import {
 import {
   loadModel,
   supportsParallelToolCallsParam,
-  Task,
 } from "../../../utils/llms/index.js";
+import { LLMTask } from "@open-swe/shared/open-swe/llm-task";
 import { z } from "zod";
 import {
   getActiveTask,
@@ -148,12 +148,15 @@ export async function updatePlan(
     ...updatePlanToolCall,
   });
 
-  const model = await loadModel(config, Task.PROGRAMMER);
+  const model = await loadModel(config, LLMTask.PROGRAMMER);
   const modelManager = getModelManager();
-  const modelName = modelManager.getModelNameForTask(config, Task.PROGRAMMER);
+  const modelName = modelManager.getModelNameForTask(
+    config,
+    LLMTask.PROGRAMMER,
+  );
   const modelSupportsParallelToolCallsParam = supportsParallelToolCallsParam(
     config,
-    Task.PROGRAMMER,
+    LLMTask.PROGRAMMER,
   );
   const modelWithTools = model.bindTools([updatePlanTool], {
     tool_choice: updatePlanTool.name,

apps/open-swe/src/graphs/reviewer/nodes/final-review.ts

@@ -17,8 +17,8 @@ import {
 import {
   loadModel,
   supportsParallelToolCallsParam,
-  Task,
 } from "../../../utils/llms/index.js";
+import { LLMTask } from "@open-swe/shared/open-swe/llm-task";
 import { GraphConfig, PlanItem } from "@open-swe/shared/open-swe/types";
 import { z } from "zod";
 import { addTaskPlanToIssue } from "../../../utils/github/issue-task.js";
@@ -117,12 +117,12 @@ export async function finalReview(
   const completedTool = createCodeReviewMarkTaskCompletedFields();
   const incompleteTool = createCodeReviewMarkTaskNotCompleteFields();
   const tools = [completedTool, incompleteTool];
-  const model = await loadModel(config, Task.REVIEWER);
+  const model = await loadModel(config, LLMTask.REVIEWER);
   const modelManager = getModelManager();
-  const modelName = modelManager.getModelNameForTask(config, Task.REVIEWER);
+  const modelName = modelManager.getModelNameForTask(config, LLMTask.REVIEWER);
   const modelSupportsParallelToolCallsParam = supportsParallelToolCallsParam(
     config,
-    Task.REVIEWER,
+    LLMTask.REVIEWER,
   );
   const modelWithTools = model.bindTools(tools, {
     tool_choice: "any",

apps/open-swe/src/graphs/reviewer/nodes/generate-review-actions/index.ts

@@ -3,8 +3,8 @@ import {
   loadModel,
   Provider,
   supportsParallelToolCallsParam,
-  Task,
 } from "../../../../utils/llms/index.js";
+import { LLMTask } from "@open-swe/shared/open-swe/llm-task";
 import {
   ReviewerGraphState,
   ReviewerGraphUpdate,
@@ -189,10 +189,10 @@ export async function generateReviewActions(
   config: GraphConfig,
 ): Promise<ReviewerGraphUpdate> {
   const modelManager = getModelManager();
-  const modelName = modelManager.getModelNameForTask(config, Task.REVIEWER);
+  const modelName = modelManager.getModelNameForTask(config, LLMTask.REVIEWER);
   const modelSupportsParallelToolCallsParam = supportsParallelToolCallsParam(
     config,
-    Task.REVIEWER,
+    LLMTask.REVIEWER,
   );
   const isAnthropicModel = modelName.includes("claude-");
 
@@ -201,7 +201,7 @@ export async function generateReviewActions(
     config,
   );
 
-  const model = await loadModel(config, Task.REVIEWER, {
+  const model = await loadModel(config, LLMTask.REVIEWER, {
     providerTools,
     providerMessages,
   });

apps/open-swe/src/graphs/shared/diagnose-error.ts

@@ -14,8 +14,8 @@ import { getMessageString } from "../../utils/message/content.js";
 import {
   loadModel,
   supportsParallelToolCallsParam,
-  Task,
 } from "../../utils/llms/index.js";
+import { LLMTask } from "@open-swe/shared/open-swe/llm-task";
 import { trackCachePerformance } from "../../utils/caching.js";
 import { getModelManager } from "../../utils/llms/model-manager.js";
 
@@ -95,12 +95,15 @@ export async function diagnoseError(
 
   logger.info("The last few tool calls resulted in errors. Diagnosing error.");
 
-  const model = await loadModel(config, Task.SUMMARIZER);
+  const model = await loadModel(config, LLMTask.SUMMARIZER);
   const modelManager = getModelManager();
-  const modelName = modelManager.getModelNameForTask(config, Task.SUMMARIZER);
+  const modelName = modelManager.getModelNameForTask(
+    config,
+    LLMTask.SUMMARIZER,
+  );
   const modelSupportsParallelToolCallsParam = supportsParallelToolCallsParam(
     config,
-    Task.SUMMARIZER,
+    LLMTask.SUMMARIZER,
   );
   const modelWithTools = model.bindTools([diagnoseErrorTool], {
     tool_choice: diagnoseErrorTool.name,

apps/open-swe/src/tools/search-documents-for/index.ts

@@ -2,7 +2,8 @@ import { tool } from "@langchain/core/tools";
 import { createLogger, LogLevel } from "../../utils/logger.js";
 import { createSearchDocumentForToolFields } from "@open-swe/shared/open-swe/tools";
 import { FireCrawlLoader } from "@langchain/community/document_loaders/web/firecrawl";
-import { loadModel, Task } from "../../utils/llms/index.js";
+import { loadModel } from "../../utils/llms/index.js";
+import { LLMTask } from "@open-swe/shared/open-swe/llm-task";
 import { GraphConfig, GraphState } from "@open-swe/shared/open-swe/types";
 import { getMessageContentString } from "@open-swe/shared/messages";
 import { DOCUMENT_SEARCH_PROMPT } from "./prompt.js";
@@ -76,7 +77,7 @@ export function createSearchDocumentForTool(
           };
         }
 
-        const model = await loadModel(config, Task.SUMMARIZER);
+        const model = await loadModel(config, LLMTask.SUMMARIZER);
 
         const searchPrompt = DOCUMENT_SEARCH_PROMPT.replace(
           "{DOCUMENT_PAGE_CONTENT}",

apps/open-swe/src/utils/llms/index.ts

@@ -1,3 +1,2 @@
 export * from "./load-model.js";
-export * from "./constants.js";
 export * from "./model-manager.js";

apps/open-swe/src/utils/llms/load-model.ts

@@ -1,13 +1,16 @@
 import { GraphConfig } from "@open-swe/shared/open-swe/types";
 import { getModelManager, Provider } from "./model-manager.js";
 import { FallbackRunnable } from "../runtime-fallback.js";
-import { Task, TASK_TO_CONFIG_DEFAULTS_MAP } from "./constants.js";
 import { BindToolsInput } from "@langchain/core/language_models/chat_models";
 import { BaseMessageLike } from "@langchain/core/messages";
+import {
+  LLMTask,
+  TASK_TO_CONFIG_DEFAULTS_MAP,
+} from "@open-swe/shared/open-swe/llm-task";
 
 export async function loadModel(
   config: GraphConfig,
-  task: Task,
+  task: LLMTask,
   options?: {
     providerTools?: Record<Provider, BindToolsInput[]>;
     providerMessages?: Record<Provider, BaseMessageLike[]>;
@@ -33,7 +36,7 @@ export const MODELS_NO_PARALLEL_TOOL_CALLING = ["openai:o3", "openai:o3-mini"];
 
 export function supportsParallelToolCallsParam(
   config: GraphConfig,
-  task: Task,
+  task: LLMTask,
 ): boolean {
   const modelStr =
     config.configurable?.[`${task}ModelName`] ??

apps/open-swe/src/utils/llms/model-manager.ts

@@ -4,10 +4,12 @@ import {
 } from "langchain/chat_models/universal";
 import { GraphConfig } from "@open-swe/shared/open-swe/types";
 import { createLogger, LogLevel } from "../logger.js";
-import { Task } from "./constants.js";
+import {
+  LLMTask,
+  TASK_TO_CONFIG_DEFAULTS_MAP,
+} from "@open-swe/shared/open-swe/llm-task";
 import { isAllowedUser } from "@open-swe/shared/github/allowed-users";
 import { decryptSecret } from "@open-swe/shared/crypto";
-import { TASK_TO_CONFIG_DEFAULTS_MAP } from "./constants.js";
 import { API_KEY_REQUIRED_MESSAGE } from "@open-swe/shared/constants";
 
 const logger = createLogger(LogLevel.INFO, "ModelManager");
@@ -101,7 +103,7 @@ export class ModelManager {
   /**
    * Load a single model (no fallback during loading)
    */
-  async loadModel(graphConfig: GraphConfig, task: Task) {
+  async loadModel(graphConfig: GraphConfig, task: LLMTask) {
     const baseConfig = this.getBaseConfigForTask(graphConfig, task);
     const model = await this.initializeModel(baseConfig, graphConfig);
     return model;
@@ -199,7 +201,7 @@ export class ModelManager {
 
   public getModelConfigs(
     config: GraphConfig,
-    task: Task,
+    task: LLMTask,
     selectedModel: ConfigurableModel,
   ) {
     const configs: ModelLoadConfig[] = [];
@@ -264,7 +266,7 @@ export class ModelManager {
   /**
    * Get the model name for a task from GraphConfig
    */
-  public getModelNameForTask(config: GraphConfig, task: Task): string {
+  public getModelNameForTask(config: GraphConfig, task: LLMTask): string {
     const baseConfig = this.getBaseConfigForTask(config, task);
     return baseConfig.modelName;
   }
@@ -274,34 +276,34 @@ export class ModelManager {
    */
   private getBaseConfigForTask(
     config: GraphConfig,
-    task: Task,
+    task: LLMTask,
   ): ModelLoadConfig {
     const taskMap = {
-      [Task.PLANNER]: {
+      [LLMTask.PLANNER]: {
         modelName:
           config.configurable?.[`${task}ModelName`] ??
           TASK_TO_CONFIG_DEFAULTS_MAP[task].modelName,
         temperature: config.configurable?.[`${task}Temperature`] ?? 0,
       },
-      [Task.PROGRAMMER]: {
+      [LLMTask.PROGRAMMER]: {
         modelName:
           config.configurable?.[`${task}ModelName`] ??
           TASK_TO_CONFIG_DEFAULTS_MAP[task].modelName,
         temperature: config.configurable?.[`${task}Temperature`] ?? 0,
       },
-      [Task.REVIEWER]: {
+      [LLMTask.REVIEWER]: {
         modelName:
           config.configurable?.[`${task}ModelName`] ??
           TASK_TO_CONFIG_DEFAULTS_MAP[task].modelName,
         temperature: config.configurable?.[`${task}Temperature`] ?? 0,
       },
-      [Task.ROUTER]: {
+      [LLMTask.ROUTER]: {
         modelName:
           config.configurable?.[`${task}ModelName`] ??
           TASK_TO_CONFIG_DEFAULTS_MAP[task].modelName,
         temperature: config.configurable?.[`${task}Temperature`] ?? 0,
       },
-      [Task.SUMMARIZER]: {
+      [LLMTask.SUMMARIZER]: {
         modelName:
           config.configurable?.[`${task}ModelName`] ??
           TASK_TO_CONFIG_DEFAULTS_MAP[task].modelName,
@@ -341,29 +343,29 @@ export class ModelManager {
    */
   private getDefaultModelForProvider(
     provider: Provider,
-    task: Task,
+    task: LLMTask,
   ): ModelLoadConfig | null {
-    const defaultModels: Record<Provider, Record<Task, string>> = {
+    const defaultModels: Record<Provider, Record<LLMTask, string>> = {
       anthropic: {
-        [Task.PLANNER]: "claude-sonnet-4-0",
-        [Task.PROGRAMMER]: "claude-sonnet-4-0",
-        [Task.REVIEWER]: "claude-sonnet-4-0",
-        [Task.ROUTER]: "claude-3-5-haiku-latest",
-        [Task.SUMMARIZER]: "claude-sonnet-4-0",
+        [LLMTask.PLANNER]: "claude-sonnet-4-0",
+        [LLMTask.PROGRAMMER]: "claude-sonnet-4-0",
+        [LLMTask.REVIEWER]: "claude-sonnet-4-0",
+        [LLMTask.ROUTER]: "claude-3-5-haiku-latest",
+        [LLMTask.SUMMARIZER]: "claude-sonnet-4-0",
       },
       "google-genai": {
-        [Task.PLANNER]: "gemini-2.5-flash",
-        [Task.PROGRAMMER]: "gemini-2.5-pro",
-        [Task.REVIEWER]: "gemini-2.5-flash",
-        [Task.ROUTER]: "gemini-2.5-flash",
-        [Task.SUMMARIZER]: "gemini-2.5-pro",
+        [LLMTask.PLANNER]: "gemini-2.5-flash",
+        [LLMTask.PROGRAMMER]: "gemini-2.5-pro",
+        [LLMTask.REVIEWER]: "gemini-2.5-flash",
+        [LLMTask.ROUTER]: "gemini-2.5-flash",
+        [LLMTask.SUMMARIZER]: "gemini-2.5-pro",
       },
       openai: {
-        [Task.PLANNER]: "o3",
-        [Task.PROGRAMMER]: "gpt-4.1",
-        [Task.REVIEWER]: "o3",
-        [Task.ROUTER]: "gpt-4o-mini",
-        [Task.SUMMARIZER]: "gpt-4.1-mini",
+        [LLMTask.PLANNER]: "o3",
+        [LLMTask.PROGRAMMER]: "gpt-4.1",
+        [LLMTask.REVIEWER]: "o3",
+        [LLMTask.ROUTER]: "gpt-4o-mini",
+        [LLMTask.SUMMARIZER]: "gpt-4.1-mini",
       },
     };
 

apps/open-swe/src/utils/mcp-output/index.ts

@@ -1,5 +1,6 @@
 import { GraphConfig } from "@open-swe/shared/open-swe/types";
-import { loadModel, Task } from "../llms/index.js";
+import { loadModel } from "../llms/index.js";
+import { LLMTask } from "@open-swe/shared/open-swe/llm-task";
 import { createLogger, LogLevel } from "../logger.js";
 import { DOCUMENT_TOC_GENERATION_PROMPT } from "./prompt.js";
 import { getMessageContentString } from "@open-swe/shared/messages";
@@ -29,7 +30,7 @@ export async function handleMcpDocumentationOutput(
   });
 
   try {
-    const model = await loadModel(config, Task.SUMMARIZER);
+    const model = await loadModel(config, LLMTask.SUMMARIZER);
 
     const systemPrompt = DOCUMENT_TOC_GENERATION_PROMPT.replace(
       "{DOCUMENT_PAGE_CONTENT}",

apps/open-swe/src/utils/runtime-fallback.ts

@@ -1,5 +1,5 @@
 import { GraphConfig } from "@open-swe/shared/open-swe/types";
-import { Task } from "./llms/index.js";
+import { LLMTask } from "@open-swe/shared/open-swe/llm-task";
 import { ModelManager, Provider } from "./llms/model-manager.js";
 import { createLogger, LogLevel } from "./logger.js";
 import { Runnable, RunnableConfig } from "@langchain/core/runnables";
@@ -45,7 +45,7 @@ export class FallbackRunnable<
 > extends ConfigurableModel<RunInput, CallOptions> {
   private primaryRunnable: any;
   private config: GraphConfig;
-  private task: Task;
+  private task: LLMTask;
   private modelManager: ModelManager;
   private providerTools?: Record<Provider, BindToolsInput[]>;
   private providerMessages?: Record<Provider, BaseMessageLike[]>;
@@ -53,7 +53,7 @@ export class FallbackRunnable<
   constructor(
     primaryRunnable: any,
     config: GraphConfig,
-    task: Task,
+    task: LLMTask,
     modelManager: ModelManager,
     options?: {
       providerTools?: Record<Provider, BindToolsInput[]>;

apps/web/src/components/api-key-banner.tsx

@@ -8,6 +8,7 @@ import { useUser } from "@/hooks/useUser";
 import { useConfigStore, DEFAULT_CONFIG_KEY } from "@/hooks/useConfigStore";
 import { isAllowedUser } from "@open-swe/shared/github/allowed-users";
 import Link from "next/link";
+import { hasApiKeySet } from "@/lib/api-keys";
 
 const API_KEY_BANNER_DISMISSED_KEY = "api_key_banner_dismissed";
 
@@ -28,29 +29,15 @@ export function ApiKeyBanner() {
     }
   }, []);
 
-  // Don't show banner if:
-  // - Still loading user data
-  // - User is not authenticated
-  // - User has dismissed the banner
-  if (isLoading || !user || dismissed) {
-    return null;
-  }
+  const userIsAllowed = user && isAllowedUser(user.login);
 
-  // Check if user is in the allowed list
-  const userIsAllowed = isAllowedUser(user.login);
-
-  // If user is allowed, they don't need API keys
-  if (userIsAllowed) {
-    return null;
-  }
-
-  // Check if user has any API keys configured
-  const apiKeys = config.apiKeys || {};
-  const hasApiKeys =
-    apiKeys.anthropicApiKey || apiKeys.openaiApiKey || apiKeys.googleApiKey;
-
-  // If user has API keys, don't show banner
-  if (hasApiKeys) {
+  if (
+    isLoading ||
+    !user ||
+    dismissed ||
+    userIsAllowed ||
+    hasApiKeySet(config)
+  ) {
     return null;
   }
 

apps/web/src/components/v2/terminal-input.tsx

@@ -20,6 +20,9 @@ import {
 } from "@open-swe/shared/constants";
 import { ManagerGraphUpdate } from "@open-swe/shared/open-swe/manager/types";
 import { useDraftStorage } from "@/hooks/useDraftStorage";
+import { hasApiKeySet } from "@/lib/api-keys";
+import { useUser } from "@/hooks/useUser";
+import { isAllowedUser } from "@open-swe/shared/github/allowed-users";
 
 interface TerminalInputProps {
   placeholder?: string;
@@ -36,6 +39,24 @@ interface TerminalInputProps {
   draftToLoad?: string;
 }
 
+const MISSING_API_KEYS_TOAST_CONTENT = (
+  <p>
+    {API_KEY_REQUIRED_MESSAGE} Please add your API key(s) in{" "}
+    <a
+      className="text-blue-500 underline underline-offset-1 dark:text-blue-400"
+      href="/settings?tab=api-keys"
+    >
+      settings
+    </a>
+  </p>
+);
+
+const MISSING_API_KEYS_TOAST_OPTIONS = {
+  richColors: true,
+  duration: 30_000,
+  closeButton: true,
+};
+
 export function TerminalInput({
   placeholder = "Enter your command...",
   disabled = false,
@@ -55,6 +76,7 @@ export function TerminalInput({
   const { getConfig } = useConfigStore();
   const { selectedRepository } = useGitHubAppProvider();
   const [loading, setLoading] = useState(false);
+  const { user, isLoading: isUserLoading } = useUser();
 
   const stream = useStream<GraphState>({
     apiUrl,
@@ -71,8 +93,28 @@ export function TerminalInput({
       });
       return;
     }
+
+    if (!user) {
+      toast.error("User not found. Please sign in first", {
+        richColors: true,
+        closeButton: true,
+      });
+      return;
+    }
+
+    const defaultConfig = getConfig(DEFAULT_CONFIG_KEY);
+
+    if (!isAllowedUser(user.login) && !hasApiKeySet(defaultConfig)) {
+      toast.error(
+        MISSING_API_KEYS_TOAST_CONTENT,
+        MISSING_API_KEYS_TOAST_OPTIONS,
+      );
+    }
+
     setLoading(true);
+
     const trimmedMessage = message.trim();
+
     if (trimmedMessage.length > 0 || contentBlocks.length > 0) {
       const newHumanMessage = new HumanMessage({
         id: uuidv4(),
@@ -91,6 +133,7 @@ export function TerminalInput({
           targetRepository: selectedRepository,
           autoAcceptPlan,
         };
+
         const run = await stream.client.runs.create(
           newThreadId,
           MANAGER_GRAPH_ID,
@@ -99,7 +142,7 @@ export function TerminalInput({
             config: {
               recursion_limit: 400,
               configurable: {
-                ...getConfig(DEFAULT_CONFIG_KEY),
+                ...defaultConfig,
               },
             },
             ifNotExists: "create",
@@ -144,20 +187,8 @@ export function TerminalInput({
           e.message.includes(API_KEY_REQUIRED_MESSAGE)
         ) {
           toast.error(
-            <p>
-              {API_KEY_REQUIRED_MESSAGE} Please add your API key(s) in{" "}
-              <a
-                className="text-blue-500 underline underline-offset-1 dark:text-blue-400"
-                href="/settings?tab=api-keys"
-              >
-                settings
-              </a>
-            </p>,
-            {
-              richColors: true,
-              duration: 30_000,
-              closeButton: true,
-            },
+            MISSING_API_KEYS_TOAST_CONTENT,
+            MISSING_API_KEYS_TOAST_OPTIONS,
           );
         }
       } finally {
@@ -205,7 +236,9 @@ export function TerminalInput({
 
         <Button
           onClick={handleSend}
-          disabled={disabled || !message.trim() || !selectedRepository}
+          disabled={
+            disabled || !message.trim() || !selectedRepository || isUserLoading
+          }
           size="icon"
           variant="brand"
           className="ml-auto size-8 rounded-full border border-white/20 transition-all duration-200 hover:border-white/30 disabled:border-transparent"

apps/web/src/lib/api-keys.ts

@@ -0,0 +1,25 @@
+export function hasApiKeySet(config: Record<string, any>) {
+  const modelNameKeys = Object.keys(config).filter((key) =>
+    key.endsWith("ModelName"),
+  );
+  const enabledProviders = modelNameKeys
+    .map((key) => config[key])
+    .map((p) => p.split(":")[0]);
+
+  const apiKeys = config.apiKeys || {};
+
+  // No providers enabled means user is using default model: anthropic
+  if (enabledProviders.length === 0 && !apiKeys.anthropicApiKey) {
+    return false;
+  }
+
+  if (
+    (enabledProviders.includes("anthropic") && !apiKeys.anthropicApiKey) ||
+    (enabledProviders.includes("openai") && !apiKeys.openaiApiKey) ||
+    (enabledProviders.includes("google-genai") && !apiKeys.googleApiKey)
+  ) {
+    return false;
+  }
+
+  return true;
+}

packages/shared/src/open-swe/llm-task.ts

@@ -1,4 +1,4 @@
-export enum Task {
+export enum LLMTask {
   /**
    * Used for programmer tasks. This includes: writing code,
    * generating plans, taking context gathering actions, etc.
@@ -28,23 +28,23 @@ export enum Task {
 }
 
 export const TASK_TO_CONFIG_DEFAULTS_MAP = {
-  [Task.PLANNER]: {
+  [LLMTask.PLANNER]: {
     modelName: "anthropic:claude-sonnet-4-0",
     temperature: 0,
   },
-  [Task.PROGRAMMER]: {
+  [LLMTask.PROGRAMMER]: {
     modelName: "anthropic:claude-sonnet-4-0",
     temperature: 0,
   },
-  [Task.REVIEWER]: {
+  [LLMTask.REVIEWER]: {
     modelName: "anthropic:claude-sonnet-4-0",
     temperature: 0,
   },
-  [Task.ROUTER]: {
+  [LLMTask.ROUTER]: {
     modelName: "anthropic:claude-3-5-haiku-latest",
     temperature: 0,
   },
-  [Task.SUMMARIZER]: {
+  [LLMTask.SUMMARIZER]: {
     modelName: "anthropic:claude-3-5-haiku-latest",
     temperature: 0,
   },