Skip to main content
Tools extend what agents can do beyond text conversations. Each tool gives agents specific capabilities to interact with the real world.

​
How Tools Work

When you make a request, the agent automatically selects the right tools. You don’t need to specify which tool to use — just describe what you want. Example: “Search for the latest news about AI regulations” → Agent uses Web Search tool automatically.

​
Available Tools

​
Search & Research

Web Search

Search the internet for current information using multiple search engines.

Browser

Navigate websites, extract content, take screenshots, and interact with web pages.

YouTube

Extract transcripts, metadata, and summaries from YouTube videos.

​
Content & Media

File Upload

Process documents (PDF, DOCX, TXT), images, and spreadsheets.

Image Generation

Create images using AI models like DALL-E.

Audio Processing

Transcribe audio files, text-to-speech, and audio analysis.

​
Productivity

Scheduler

Create, modify, and manage scheduled tasks and recurring automations.

Memory

Store and recall important information across conversations.

​
Developer

Code Execution

Run code snippets and scripts in sandboxed environments.

Git

Interact with Git repositories — commits, branches, PRs.

​
Tool Selection

Agents select tools based on:
  1. Your request — What you’re asking for
  2. Available tools — Which tools are enabled for the agent
  3. Connected services — Which external services are configured
You can limit which tools an agent has access to in the agent’s settings. This is useful for creating focused agents that only handle specific tasks.

​
Tool Results

Tool results appear inline in the conversation. Depending on the tool:
  • Search results show as summarized findings with source links
  • Generated images appear as inline images
  • File processing shows extracted content or analysis
  • Browser can show screenshots or extracted data