Files
BrowserOS/reference-code/old-lib/agent/ProductivityAgent.ts
Felarof 8245dfe0ff Rewrite Agent Loop (#7)
* clean-up bunch of files for re-write

* more clean-up and adding basic agent

* Minor fix moved types into respective files.

* Deleted bunch of old files

backup

Update gitignore

Deleted a bunch of files

Remove message manager

Deleted old docs

Update rules

rename Profiler to profiler

* Temporarily adding old code

* Adding two small things back

* backup

* Implemented LangChainProvider and updated cursor rules

backup

LangChainProvider

curosr rules

* Implement tests for LangChainProvider -- unit test and integration test

integration test passes

integration test backup

* Tool Design

Tools Desing

tools design

* NavigationTool ready

NavigationTool ready

NavigationTool ready

NaivgationTool ready

backup

* MessageManager

MessageManager

backup

* Fixed integration test

* Agent design new

Updated agent design and added bunch of /NTN commands

agent new design

* Delete old agent design

* MessageManagerReadOnly class

* PlannerTool ready

PlannerTool almost ready

* ToolManager and DoneTool

* Integration of BrowserAgent

* BrowserAgent implementation v0.1

* BrowserAgent small fix v0.2

* Tool calling design

too call design

tool design claude

* Update agent tool design with // NTN

* add zod-to-json npm install

* BrowserAGent v0.3

* BrowserAgent v0.4

* BrowserAgent v0.5

* fixes

* Build error fixes in my NEWLY added code

build errors fix

* Build error fixes in old code (integration work)

backup

* Comment StreamEventProcessor for now, it is not used

* Small build error fix

* Small rename

* Added integration test to check structuredLLM and changed to 4o-mini

change default to nxtscape

integration test

* Small docstring

* Simplified BrowserAgent code and added integration test

Simplified BrowserAgent code

BrowserAGent integrationt est

* Update CLAUDE.md with project memory and instructions on how to write code

Update CLAUDE.md with project memory and instructions on how to write code

Project Memory

* Just a mova.. Moved ToolManager outside. Build works.

* TabOperations tool

TabOperations Tool and fixing some test

tab operations

* Update CLAUDE.md

* Added ClassificationTool

classifiction tool

classification prommpt

* Refactored and simplified PlannerTool unit test and integration test

* Updated Plnnaer tool

* Update CLAUDE.md

* BrowserAgent modified to do classification

BrowserAgent with classification

* minor fix to ToolManager

* Instead of ToolCall and ToolResult -- just updating message manager once

* minor fix to BrowserAgent integration test

* Changed done to "done_tool"

* Updated CLAUDE.md to reflect understanding of claude

* Uncommented stream event processor

* Renamed EventBus to StreamEventBus

* Commented StreamEventProcessor

* Event Processor

* Integrated EventProcessor with BrowserAgent

Added EventProcessor to BrowserAgetn

* Renamed StreamEventBus to EventBus

* Made EventBus required parameter in ExecutionContext

* PlanGenerator rewrite

PlanGenerator rewrite

backup

* For simple task, explicitly tell it to call done tool

* Max attempts for simple task

* backup

* Revert "backup"

This reverts commit 7d79a3d4d5774bfef79ec9827878b74edad3593f.

* Consolidating where EventBus and EventProcessor are created and initialized

backup

* Update CLAUDE.md

Update CLAUDE.md

* Improving agent loop code

Cleaned up processTooCall

classification task

* Create test-writer subAgent

test-agent-prompt

test agent prompt

test-agent-prompt

Update test-writer.md

* BrowserAgent test

Browseragent test

BrowserAgent test

* BrowserAgent refactor

backup

backup

* Minor fixes

* Minor fix

* minor change -- NEW AGENT LOOP IS WORKING WELL

* Update cursor rules

* Small change

* Improved BrowserAgent integration test

Improved BrowserAgent integration test

* Small change

* Update CLAUDE.md

* Different tools

* FindElementTool is ready

Find element update

backup

find element backup

* Updated to test strings to say "tests..."

* ScrollTool is ready

* RefreshStateTool is updated as well

* MessageManager updated

* SearchTool is ready

backup

* Interaction Element is also ready

* Add debugMessage emitter

* ValidatorTool ready and tests are passing

Validation Tool

validator tool

backup

backup

* GroupTabs tool ready

* Registered all the tools

* Planning changed to 5 steps

* BrowserAgent integration test fix

* Minor string changes

* backup

* Removed too many confusing events in EventProcessor -- there is only event.info right now

* Abort control implemented

backup

Abort

* Formatter for toolResult

Formatter for toolResult

backup

* Always render using Markdown

* Minor fix

---------

Co-authored-by: Nikhil Sonti <nikhilsv92@gmail.com>
2025-07-29 08:14:45 -07:00

201 lines
6.5 KiB
TypeScript

import { z } from 'zod';
import { createReactAgent } from '@langchain/langgraph/prebuilt';
import { Logging } from '@/lib/utils/Logging';
import { RunnableConfig } from '@langchain/core/runnables';
// Import base agent
import { BaseAgent, AgentOptions, AgentInput } from './BaseAgent';
// Import new prompt system
import { ProductivityAgentPrompt } from '@/lib/prompts';
// Import new tool system
import { ToolRegistry } from '@/lib/tools/base';
import {
TabOperationsTool,
GroupTabsTool
} from '@/lib/tools/tab';
import {
SaveBookmarkTool,
BookmarkManagementTool,
BookmarkSearchTool,
BookmarksFolderTool
} from '@/lib/tools/bookmarks';
import { SessionManagementTool, SessionExecutionTool } from '@/lib/tools/sessions';
import { NoOpTool, TerminateTool, GetDateTool } from '@/lib/tools/utility';
import { GetSelectedTabsTool } from '@/lib/tools/tab';
import { GetHistoryTool, StatsHistoryTool } from '@/lib/tools/history';
/**
* Productivity agent output schema
*/
export const ProductivityOutputSchema = z.object({
completed: z.boolean(), // Whether the productivity task was completed
result: z.string(), // Description of what was accomplished
data: z.record(z.unknown()).optional() // Any data retrieved (tab info, bookmarks, etc.)
});
export type ProductivityOutput = z.infer<typeof ProductivityOutputSchema>;
/**
* Agent specialized for productivity features like tab management, workspace organization,
* and browser efficiency improvements.
*/
export class ProductivityAgent extends BaseAgent {
/**
* Creates a new instance of ProductivityAgent
* @param options - Configuration options for the productivity agent
*/
constructor(options: AgentOptions) {
super(options);
}
/**
* Override: Create tool registry for the agent
* @returns ToolRegistry with productivity tools
*/
protected createToolRegistry(): ToolRegistry {
Logging.log(
"ProductivityAgent",
"🔧 Creating ToolRegistry with ALL productivity tools (including TabOperationsTool)",
"info"
);
const registry = new ToolRegistry();
// Register productivity tools
registry.registerAll([
new TabOperationsTool(this.executionContext),
new GroupTabsTool(this.executionContext),
new GetSelectedTabsTool(this.executionContext),
new NoOpTool(this.executionContext),
new TerminateTool(this.executionContext),
new GetDateTool(this.executionContext),
new SessionManagementTool(this.executionContext),
new SessionExecutionTool(this.executionContext),
new SaveBookmarkTool(this.executionContext),
new BookmarkManagementTool(this.executionContext),
new BookmarkSearchTool(this.executionContext),
new BookmarksFolderTool(this.executionContext),
new GetHistoryTool(this.executionContext),
new StatsHistoryTool(this.executionContext),
]);
Logging.log(
"ProductivityAgent",
"🔧 Tools registered:" +
registry
.getAll()
.map((t: any) => t.getConfig().name)
.join(", "),
"info"
);
return registry;
}
/**
* Override: Generate system prompt for productivity agent
* @returns System prompt string
*/
protected generateSystemPrompt(): string {
// Use the tool registry to generate documentation
const toolDocs = this.toolRegistry?.generateSystemPrompt() || "";
// Create and use the new prompt generator
const promptGenerator = new ProductivityAgentPrompt(toolDocs);
return promptGenerator.generate();
}
/**
* Override: Get the agent name for logging
* @returns Agent name
*/
protected getAgentName(): string {
return "ProductivityAgent";
}
/**
* Execute productivity agent - handles instruction enhancement and execution
* @param input - Agent input containing instruction and context
* @param callbacks - Optional streaming callbacks
* @param config - Optional configuration for LangGraph web compatibility
* @returns Parsed productivity output
*/
protected async executeAgent(
input: AgentInput,
config?: RunnableConfig
): Promise<ProductivityOutput> {
try {
await this.ensureInitialized();
// Create the ReAct agent
this.log(`🎯 Creating productivity agent`);
// 1. Add system prompt to message history at position 0 (agent-specific)
this.executionContext.messageManager.addSystemMessage(this.systemPrompt, 0);
this.systemPromptAdded = true;
const selectedTabsInstruction = await this.getSelectedTabsInstruction();
if (selectedTabsInstruction) {
this.executionContext.messageManager.addHumanMessage(`[Context: ${selectedTabsInstruction}]`);
}
// Get LLM and tools
const llm = await this.getLLM();
const tools = this.createTools();
const isGemini = llm._llmType()?.indexOf('google') !== -1 || false;
const messages = this.executionContext.messageManager.getMessages(isGemini);
// Create ReAct agent
const agent = createReactAgent({
llm,
tools,
});
// Use centralized streaming execution
const { result, allMessages } = await this.executeReactAgentWithStreaming(
agent,
input.instruction,
config,
messages
);
// TODO(nithin): This final message is often just some AImessagechunk from langchain and it has nothing.
// I should remove extracting this final message or using it for anything like marking done or something.
const finalMessage = allMessages[allMessages.length - 1];
const resultText =
typeof finalMessage?.content === "string"
? finalMessage.content
: "Task completed";
// Remove system prompt after execution
if (this.systemPromptAdded) {
this.executionContext.messageManager.removeSystemMessage();
this.systemPromptAdded = false;
}
this.log(`✅ Productivity task completed successfully`);
return {
completed: true,
result: resultText,
data: {}, // Can be enhanced to include specific data based on task type
};
} catch (error) {
// Ensure system prompt is cleaned up on error
if (this.systemPromptAdded) {
this.executionContext.messageManager.removeSystemMessage();
this.systemPromptAdded = false;
}
const errorMessage =
error instanceof Error ? error.message : String(error);
this.log(`❌ Productivity task failed: ${errorMessage}`, "error");
return {
completed: false,
result: `Task failed: ${errorMessage}`,
data: {},
};
}
}
}