# Toolhouse
## Welcome to Toolhouse
Toolhouse allows everyone to create AI workers — specialized agents that are good at performing repetitive work autonomously and reliably.
You describe what you want your worker to do in plain language. Toolhouse handles the rest: writing the prompt, selecting the right integrations, and giving your worker the tools it needs to get the job done. No coding required.
***
### What you can build
Toolhouse workers are well-suited for any task that is repetitive, time-consuming, or spans multiple tools. Common examples include:
* **Executive Assistant:** manages your calendar, triages your inbox, summarizes what needs your attention, and drafts responses on your behalf. Keeps you focused on the work only you can do.
* **Legal Counsel:** reviews contracts and documents against your guidelines, flags clauses that need attention, and drafts standard agreements using your templates. Available instantly, at any hour.
* **Marketing Team:** plans and schedules campaigns across your email platform, generates copy and visuals using your brand guidelines, and reports on performance — without waiting for a brief to be picked up.
* **Social Media Manager:** manages your social media presence end-to-end. Drafts posts, applies your brand voice, schedules content, and monitors engagement — all without manual intervention.
* **Customer Support Worker**: reads your documentation and support guidelines to answer incoming tickets, automatically files and resolves common issues by connecting to your ticketing platform, and notifies your team when users report bugs or patterns worth escalating.
If you find yourself doing the same multi-step task more than once a week, it's definitely a good candidate for a Toolhouse worker.
***
### Core concepts
**Agent Editor** is where you create and edit your workers. Describe your task in plain language, and Agent Editor will generate a structured prompt, ask follow-up questions, and configure the right integrations automatically. You can refine the worker at any time by telling the editor what to change.
**Integrations** give your workers access to external services and capabilities. Toolhouse adds integrations automatically based on what your worker needs, and you can also add them manually. Your workers can share integrations — if a worker is connected to Google Drive or Slack, the Virtual Computer it uses will have access to those services too.
**Agent Files** let you attach knowledge to your worker — templates, guidelines, skill definitions, reference documents. File contents are added directly to your worker's instructions, so it can draw on them at any point during a task. See [Skills & Knowledge](/toolhouse/capabilites/skills-and-knowledge.md) for more.
***
### Built-in capabilities
| Capability | What it does |
| --------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------ |
| [Scraping](/toolhouse/capabilites/scraping.md) | Reads and understands what's inside a webpage |
| [Web and Social Media Search](/toolhouse/capabilites/web-and-social-media-search.md) | Searches the web using specified queries. Can search Linkedin and X out of the box |
| [Memory](/toolhouse/capabilites/memory.md) | Memorize data, behavior, and past outputs |
| [Image generation](/toolhouse/capabilites/image-generation.md) and [Image editing](/toolhouse/capabilites/image-editing.md) | Creates images, quote cards, diagrams, and other visual assets |
| [Vision](/toolhouse/capabilites/vision.md) | Sees what's inside an image, a screenshot, a page, or a document |
| [Virtual computer](/toolhouse/capabilites/virtual-computer.md) | Runs Python code in an isolated, internet-connected sandbox for computation-heavy tasks, batch operations, and third-party API calls |
| [File download](/toolhouse/capabilites/file-download.md) | Fetches a file from a URL and makes its contents available to your worker |
| [Document parser](/toolhouse/capabilites/document-parser.md) | Fetches a file or web page from a URL and converts it to clean Markdown for easier reading |
| [Skills & Knowledge](/toolhouse/capabilites/skills-and-knowledge.md) | Attaches files to your worker: templates, guidelines, skill definitions, and reference material |
***
### How Toolhouse adds capabilities
Most capabilities are added automatically. When you describe a task that requires fetching a file, Toolhouse will attach File Download. When you describe a task that involves running analysis, it will attach the Virtual Computer. You don't need to know which integration to use before you start.
You can also add any integration manually from the Integrations section inside your worker's settings. This is useful when you want to ensure a capability is always available regardless of what you ask the worker to do.
***
### Where to go next
* **New to Toolhouse?** Start by creating your first worker in [Agent Editor](https://toolhouse.app/). Describe a task you do regularly and see what it generates.
* **Ready to get more out of your workers?** Read [Techniques and best practices](/toolhouse/toolhouse/techniques-and-best-practices.md) for tips on prompting, structuring files, and handling edge cases.
* **Need to connect to a tool Toolhouse doesn't support natively?** See [Virtual computer](/toolhouse/capabilites/virtual-computer.md#connecting-to-third-party-apis) to learn how to upload API documentation and have your worker figure out the rest.
### Jump right in
---
# Agent Instructions: Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.
Perform an HTTP GET request on the current page URL with the `ask` query parameter:
```
GET https://docs.toolhouse.ai/toolhouse/readme.md?ask=
```
The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.