> ## Documentation Index
> Fetch the complete documentation index at: https://docs.replit.com/llms.txt
> Use this file to discover all available pages before exploring further.

# General Agent

> Learn how General Agent lets you work with any framework, create any output type, and read and write to connected services.

## What is General Agent?

General Agent is the most flexible way to work with Replit Agent. You don't need to pre-select an artifact type or commit to building something specific. Just start chatting — create a CSV, generate a PDF, research a topic, or build a full app when you're ready.

Key concepts:

* **No pre-selection**: Start working without choosing an artifact type upfront
* **Any output**: Generate files (CSV, PDF, docs), do research, or build apps
* **Connectors**: Read and write to external services like BigQuery, Linear, Slack, Notion, Amplitude, Segment, Hex, and more
* **Progressive disclosure**: Start with a lightweight chat and escalate to a full app build naturally

## How to access General Agent

You can access General Agent in several ways:

1. **Select "General" before writing a prompt** on the Replit home page
2. **Open any existing project** and start chatting with Agent directly
3. **Choose any [Developer Framework](https://replit.com/templates)** — all Developer Frameworks automatically include General Agent
4. **[Import](/core-concepts/project-editor/app-setup/import) from GitHub** — Projects imported from GitHub automatically get General Agent

## How to use General Agent

### Just start chatting

Open a project and start talking. You don't need to commit to building anything specific — Agent works within your current project context and can handle whatever you throw at it.

**Example prompts:**

* "Summarize the key points from this document"
* "Create a CSV of the top 100 companies by revenue"
* "What are my top five Linear tickets?"
* "Set up a Vue.js project with TypeScript"

Agent can work with files created during the session, generate single-file outputs, and execute code as part of its reasoning — all before you decide to build anything.

### Connecting external services

Connect your tools so Agent can read and write data directly:

1. Go to **Connectors** in your project settings
2. Connect your services (BigQuery, Linear, Slack, Notion, Amplitude, Segment, Hex, and more)
3. Ask Agent to read or write data through them

**Example prompts with connectors:**

* "What are the open tickets in Linear?"
* "Summarize today's Slack messages in #engineering"
* "Pull last month's sales data from BigQuery"
* "Show me upcoming events from my calendar"
* "Create a new ticket in Linear for the onboarding bug"
* "Send a summary to #engineering in Slack"

### From chat to full app

When you're ready to build, just say so. Agent transitions from lightweight chat into building your app seamlessly — no need to start over or switch contexts.

* "Turn this into a web app"
* "Build an app based on this data"
* "Create a dashboard for these metrics"
* "Set up a Rust CLI tool"

You can start casually — research a topic, generate a report, explore some data — and escalate to a full application build when the moment is right.

## What you can build

* **Knowledge work**: Research, summarize, and analyze without building an app
* **Single-file outputs**: CSVs, PDFs, documents, data exports, and formatted reports
* **Data applications**: Query connected services and visualize results in dashboards
* **Any framework or language**: Angular, Vue, Svelte, Rust, Go, C#, Python tkinter, Godot games, CLI tools, and more
* **Full-stack apps**: When you're ready, build a complete application from your chat context

## What to expect

General Agent offers more power and flexibility, which comes with some variability:

* **Single-shot setup**: Sometimes Agent configures everything in one go
* **Iterative collaboration**: Complex setups may require multiple rounds of refinement
* **Technology limitations**: Some technologies may not be fully supported in the Nix environment

<Note>
  Unlike standard Agent with pre-configured environments, General Agent sets up its own environment and run commands based on your request. You may need to ask it to configure publishing when you're ready to deploy.
</Note>

## Frequently asked questions

<Accordion title="Can the agent see my other projects?">
  No. Each chat is contained within a single project. You can't have Project A and Project B talk to each other.
</Accordion>

<Accordion title="Why can't it access my external tool yet?">
  You need to connect your services first through Connectors. Once connected, Agent can read and write to them, but it doesn't have access by default.
</Accordion>

<Accordion title="What connectors are available?">
  Currently supported: BigQuery, Linear, Slack, Notion, Amplitude, Segment, Hex, and more. Additional connectors are being added regularly.
</Accordion>

<Accordion title="Do I need connectors to use General Agent?">
  No. Connectors are optional. You can use General Agent for knowledge work, file generation, and building apps without any external connections.
</Accordion>

## Availability

| Capability      | Core | Pro |
| --------------- | ---- | --- |
| General Agent   | ✅    | ✅   |
| Any framework   | ✅    | ✅   |
| Knowledge work  | ✅    | ✅   |
| File generation | ✅    | ✅   |
| Connectors      | ✅    | ✅   |
| Build apps      | ✅    | ✅   |

General Agent is available to all users.