Editing Your Docs
Every Docsio project includes an AI agent that can modify any part of your documentation site. Instead of manually editing files, you describe the change you want in plain language, and the agent handles the rest.
Using the AI Agent Chat
The agent panel is on the right side of the editor. Type what you want changed into the chat input and press Enter. The agent reads the relevant files, makes the edits, and writes the updated content back to your project.
Common Requests
Here are examples of what you can ask the agent to do:
| What you want | Example prompt |
|---|---|
| Change a color | "Change the primary color to blue" |
| Add a page | "Add a new page about deployment in the Guides section" |
| Rewrite content | "Rewrite the intro to be more concise" |
| Add code examples | "Add a code example showing installation with npm" |
| Update navigation | "Add a link to our GitHub repo in the navbar" |
| Edit CSS | "Make the sidebar font larger and add more spacing" |
You can ask for anything -- content changes, styling updates, structural edits, configuration tweaks, and more.
How the Agent Works
When you send a message, the agent follows a predictable pattern:
- Reads the relevant file or scans the project
- Processes your request and determines the changes
- Writes the updated content back to the project
You see inline status indicators as the agent works:
- "Reading styles..." -- the agent is examining the current CSS
- "Writing styles..." -- the agent is saving updated styles
- "Scanning project files..." -- the agent is reviewing your project structure
- "Searching for..." -- the agent is looking for specific content
The agent is aware of which page you currently have selected. When you ask it to make a change, it applies the edit to the active page by default. If you want to change a different page, mention it by name.
Live Preview
Changes appear in the live preview panel immediately after the agent writes them. There is no manual refresh or rebuild step -- the preview updates in real time as the agent edits files.
Uploading Files
You can upload files to the agent by clicking the attachment icon (paperclip) in the chat input or by selecting a file from your system.
Supported Upload Types
| File type | What happens |
|---|---|
| Images (PNG, JPG, GIF, SVG, WebP) | Placed in the /img/ directory of your project. The agent provides the Markdown syntax to embed it. |
| Documents (PDF, DOC, DOCX, TXT, MD) | Text is extracted from the file and provided to the agent as context. Useful for feeding existing content into your docs. |
After uploading an image, the agent will suggest the Markdown syntax to use it:
Document uploads (PDF, DOC) are used for context only -- the text content is extracted and sent to the agent so it can reference or incorporate the information into your documentation.
Tips for Effective Editing
- Be specific. "Make the introduction shorter" works, but "Rewrite the introduction to be 2-3 sentences focusing on what the product does" works better.
- One change at a time. The agent handles complex requests well, but breaking large changes into steps gives you more control over the result.
- Reference pages by name. If you want to edit a page other than the one currently selected, tell the agent: "On the Installation page, add a prerequisites section."
- Iterate freely. If the first result is not quite right, follow up with refinements: "Make that paragraph less technical" or "Add a note about Windows compatibility."