Delete execute_prompt tool (#1174)

* Delete execute_prompt tool
Rename execute_task to execute_prompt, add a real setting

* Clean up

* Not advanced, change default

* fix

Author: roblourens

package.json

@@ -166,10 +166,10 @@
 				}
 			},
 			{
-				"name": "execute_task",
-				"toolReferenceName": "executeTask",
-				"displayName": "Execute Task",
-				"when": "config.github.copilot.chat.advanced.taskTools.enabled",
+				"name": "execute_prompt",
+				"toolReferenceName": "executePrompt",
+				"displayName": "Execute Prompt",
+				"when": "config.github.copilot.chat.executePrompt.enabled",
 				"canBeReferencedInPrompt": true,
 				"modelDescription": "Launch a new agent to handle complex, multi-step tasks autonomously. This tool is good at researching complex questions, searching for code, and executing multi-step tasks. When you are searching for a keyword or file and are not confident that you will find the right match in the first few tries, use this agent to perform the search for you.\n\n- When the agent is done, it will return a single message back to you. The result returned by the agent is not visible to the user. To show the user the result, you should send a text message back to the user with a concise summary of the result.\n - Each agent invocation is stateless. You will not be able to send additional messages to the agent, nor will the agent be able to communicate with you outside of its final report. Therefore, your prompt should contain a highly detailed task description for the agent to perform autonomously and you should specify exactly what information the agent should return back to you in its final and only message to you.\n - The agent's outputs should generally be trusted\n - Clearly tell the agent whether you expect it to write code or just to do research (search, file reads, web fetches, etc.), since it is not aware of the user's intent",
 				"tags": [],
@@ -191,27 +191,6 @@
 					]
 				}
 			},
-			{
-				"name": "execute_prompt",
-				"toolReferenceName": "executePrompt",
-				"displayName": "Execute Prompt",
-				"when": "config.github.copilot.chat.advanced.taskTools.enabled",
-				"canBeReferencedInPrompt": true,
-				"modelDescription": "This tool can take a path to a user's prompt file as input, and execute it autonomously. If the user's prompt includes multiple references to .prompt.md files, then you should use this tool to execute those prompts in sequence.\n\n- When the agent is done, it will return a single message back to you. The result returned by the agent is not visible to the user. To show the user the result, you should send a text message back to the user with a concise summary of the result.\n - The agent's outputs should generally be trusted",
-				"tags": [],
-				"inputSchema": {
-					"type": "object",
-					"properties": {
-						"filePath": {
-							"type": "string",
-							"description": "The absolute path to the prompt file to execute"
-						}
-					},
-					"required": [
-						"filePath"
-					]
-				}
-			},
 			{
 				"name": "copilot_searchWorkspaceSymbols",
 				"toolReferenceName": "symbols",
@@ -3106,6 +3085,14 @@
 							"off",
 							"detailed"
 						]
+					},
+					"github.copilot.chat.executePrompt.enabled": {
+						"type": "boolean",
+						"default": false,
+						"markdownDescription": "%github.copilot.config.executePrompt.enabled%",
+						"tags": [
+							"experimental"
+						]
 					}
 				}
 			}

package.nls.json

@@ -300,5 +300,6 @@
 	"copilot.toolSet.runTasks.description": "Run tasks in your workspace",
 	"github.copilot.config.useResponsesApi": "Use the Responses API instead of the Chat Completions API when supported. Enables reasoning and reasoning summaries.\n\n**Note**: This is an experimental feature that is not yet activated for all users.",
 	"github.copilot.config.responsesApiReasoningEffort": "Sets the reasoning effort used for the Responses API. Requires `#github.copilot.chat.useResponsesApi#`.",
-	"github.copilot.config.responsesApiReasoningSummary": "Sets the reasoning summary style used for the Responses API. Requires `#github.copilot.chat.useResponsesApi#`."
+	"github.copilot.config.responsesApiReasoningSummary": "Sets the reasoning summary style used for the Responses API. Requires `#github.copilot.chat.useResponsesApi#`.",
+	"github.copilot.config.executePrompt.enabled": "The executePrompt tool enables the agent to execute tasks in a separate, isolated context."
 }

src/extension/prompt/node/executePromptToolCalling.ts

@@ -84,7 +84,7 @@ export class ExecutePromptToolCallingLoop extends ToolCallingLoop<IExecutePrompt
 	}
 
 	protected async getAvailableTools(): Promise<LanguageModelToolInformation[]> {
-		const excludedTools = new Set([ToolName.ExecutePrompt, ToolName.ExecuteTask, ToolName.CoreManageTodoList]);
+		const excludedTools = new Set([ToolName.ExecutePrompt, ToolName.CoreManageTodoList]);
 		return (await getAgentTools(this.instantiationService, this.options.request))
 			.filter(tool => !excludedTools.has(tool.name as ToolName))
 			// TODO can't do virtual tools at this level

src/extension/tools/common/toolNames.ts

@@ -54,7 +54,6 @@ export enum ToolName {
 	ToolReplay = 'tool_replay',
 	EditFilesPlaceholder = 'edit_files',
 	ExecutePrompt = 'execute_prompt',
-	ExecuteTask = 'execute_task',
 	CoreConfirmationTool = 'vscode_get_confirmation'
 }
 
@@ -100,7 +99,6 @@ export enum ContributedToolName {
 	ToolReplay = 'copilot_toolReplay',
 	EditFilesPlaceholder = 'copilot_editFiles',
 	ExecutePrompt = 'execute_prompt',
-	ExecuteTask = 'execute_task',
 }
 
 export const byokEditToolNamesToToolNames = {

src/extension/tools/node/allTools.ts

@@ -10,7 +10,6 @@ import './createFileTool';
 import './docTool';
 import './editNotebookTool';
 import './executePromptTool';
-import './executeTaskTool';
 import './findFilesTool';
 import './findTestsFilesTool';
 import './findTextInFilesTool';

src/extension/tools/node/executePromptTool.ts

@@ -3,22 +3,19 @@
  *  Licensed under the MIT License. See License.txt in the project root for license information.
  *--------------------------------------------------------------------------------------------*/
 
-import * as l10n from '@vscode/l10n';
 import type * as vscode from 'vscode';
-import { IFileSystemService } from '../../../platform/filesystem/common/fileSystemService';
-import { IPromptPathRepresentationService } from '../../../platform/prompts/common/promptPathRepresentationService';
 import { ChatResponseStreamImpl } from '../../../util/common/chatResponseStreamImpl';
 import { IInstantiationService } from '../../../util/vs/platform/instantiation/common/instantiation';
-import { ChatResponseMarkdownPart, ExtendedLanguageModelToolResult, LanguageModelTextPart, MarkdownString } from '../../../vscodeTypes';
+import { ChatPrepareToolInvocationPart, ExtendedLanguageModelToolResult, LanguageModelTextPart } from '../../../vscodeTypes';
 import { Conversation, Turn } from '../../prompt/common/conversation';
 import { IBuildPromptContext } from '../../prompt/common/intents';
 import { ExecutePromptToolCallingLoop } from '../../prompt/node/executePromptToolCalling';
 import { ToolName } from '../common/toolNames';
 import { CopilotToolMode, ICopilotTool, ToolRegistry } from '../common/toolsRegistry';
-import { assertFileOkForTool, formatUriForFileWidget, resolveToolInputPath } from './toolUtils';
 
 export interface IExecutePromptParams {
-	filePath: string;
+	prompt: string;
+	description: string;
 }
 
 class ExecutePromptTool implements ICopilotTool<IExecutePromptParams> {
@@ -27,41 +24,24 @@ class ExecutePromptTool implements ICopilotTool<IExecutePromptParams> {
 
 	constructor(
 		@IInstantiationService private readonly instantiationService: IInstantiationService,
-		@IFileSystemService private readonly fileSystemService: IFileSystemService,
-		@IPromptPathRepresentationService private readonly promptPathRepresentationService: IPromptPathRepresentationService,
 	) { }
 
 	async invoke(options: vscode.LanguageModelToolInvocationOptions<IExecutePromptParams>, token: vscode.CancellationToken) {
-		if (!options.input.filePath) {
-			throw new Error('Invalid input');
-		}
-
-		// Read the prompt file as text and include a reference
-		const uri = resolveToolInputPath(options.input.filePath, this.promptPathRepresentationService);
-		await this.instantiationService.invokeFunction(accessor => assertFileOkForTool(accessor, uri));
-		const promptText = (await this.fileSystemService.readFile(uri)).toString();
 
 		const loop = this.instantiationService.createInstance(ExecutePromptToolCallingLoop, {
-			toolCallLimit: 5,
-			conversation: new Conversation('', [new Turn('', { type: 'user', message: promptText })]),
-			request: {
-				...this._inputContext!.request!,
-				references: [],
-				prompt: promptText,
-				toolReferences: [],
-				modeInstructions2: { content: '' },
-				editedFileEvents: []
-			},
+			toolCallLimit: 25,
+			conversation: new Conversation('', [new Turn('', { type: 'user', message: options.input.prompt })]),
+			request: this._inputContext!.request!,
 			location: this._inputContext!.request!.location,
-			promptText,
+			promptText: options.input.prompt,
 		});
 
-		// TODO This also prevents codeblock pills from being rendered
-		// I want to render this content as thinking blocks but couldn't get it to work
+		// I want to render this content as thinking blocks when we they include tool calls
 		const stream = this._inputContext?.stream && ChatResponseStreamImpl.filter(
 			this._inputContext.stream,
-			part => !(part instanceof ChatResponseMarkdownPart)
+			part => part instanceof ChatPrepareToolInvocationPart
 		);
+
 		const loopResult = await loop.run(stream, token);
 		// Return the text of the last assistant response from the tool calling loop
 		const lastRoundResponse = loopResult.toolCallRounds.at(-1)?.response ?? loopResult.round.response ?? '';
@@ -71,14 +51,9 @@ class ExecutePromptTool implements ICopilotTool<IExecutePromptParams> {
 
 	prepareInvocation(options: vscode.LanguageModelToolInvocationPrepareOptions<IExecutePromptParams>, token: vscode.CancellationToken): vscode.ProviderResult<vscode.PreparedToolInvocation> {
 		const { input } = options;
-		if (!input.filePath) {
-			return;
-		}
 		try {
-			const uri = resolveToolInputPath(input.filePath, this.promptPathRepresentationService);
 			return {
-				invocationMessage: new MarkdownString(l10n.t`Executing prompt file ${formatUriForFileWidget(uri)}`),
-				pastTenseMessage: new MarkdownString(l10n.t`Executed prompt file ${formatUriForFileWidget(uri)}`),
+				invocationMessage: input.description,
 			};
 		} catch {
 			return;

src/extension/tools/node/executeTaskTool.ts

@@ -737,7 +737,6 @@ export namespace ConfigKey {
 
 		export const EnableClaudeCodeAgent = defineSetting<boolean | string | undefined>('chat.advanced.claudeCode.enabled', false);
 		export const ClaudeCodeDebugEnabled = defineSetting<boolean>('chat.advanced.claudeCode.debug', false);
-		export const TaskToolsEnabled = defineSetting<boolean>('chat.advanced.taskTools.enabled', true);
 		export const Gpt5AlternativePatch = defineExpSetting<boolean>('chat.advanced.gpt5AlternativePatch', false);
 	}
 
@@ -805,6 +804,7 @@ export namespace ConfigKey {
 	export const Gpt5AlternatePrompt = defineExpSetting<string>('chat.gpt5AlternatePrompt', 'default');
 	export const Gpt5CodexAlternatePrompt = defineExpSetting<'default' | 'codex'>('chat.gpt5CodexAlternatePrompt', 'default');
 	export const GrokCodeAlternatePrompt = defineExpSetting<string>('chat.grokCodeAlternatePrompt', 'default');
+	export const ExecutePromptEnabled = defineSetting<boolean>('chat.executePrompt.enabled', false);
 }
 
 export function getAllConfigKeys(): string[] {