Skip to main content
idapt
HomeCodeAI ModelsPricing
Sign inStart free
  • Home
  • Pricing
  • AI Models
  • Image models
  • Voice models
  • Video models
  • Rankings
  • New models
  • Model status
  • Multi-Model Chat
  • Voice Mode
  • Voice HUD
  • Web Search
  • Image Generation
  • Video Generation
  • Audio Generation
  • Transcription
  • Drive
  • Credentials
  • Sharing
  • Workspaces
  • Tasks
  • Memory
  • Agents
  • Subagents
  • Automations
  • Skills
  • idapt Code
  • Code Execution
  • Computers
  • Computer Use
  • Computer Assist · Soon
  • Containers · Soon
  • Cloud Computers
  • Local AI
  • AI Gateway
  • API & SDK
  • CLI
  • MCP
  • Tunnels
  • All features →
  • LLM cost calculator
  • Token counter
  • Context window checker
  • Can I run it
  • Model picker quiz
  • Savings finder
  • Video cost estimator
  • Text to speech cost
  • Transcription cost
  • API endpoint tester
  • All free tools →
  • Blog
  • Use cases
  • Comparisons
  • Best of
  • Skills
  • Learn
  • Changelog
  • Help center
  • FAQ
  • Privacy
  • Compare all models
  • Support
  • idapt Code
  • Developers
  • Quickstarts
  • API reference
  • API pricing
  • CLI
  • MCP
  • Downloads
  • Desktop
  • Badges and embeds
© idapt[email protected]TermsPrivacy PolicyLegal noticeReport content
X (Twitter)
Help Center
💬

Using @mentions and references

Found this helpful? Share it:

Give the assistant exactly the context it needs by referencing your work right in a message. Type @to pull up the picker and attach whatever helps, and type /to run a skill 💬

Reference anything with @

Type @in the message box to open the picker, then search by name. You can reference:

  • Files and folders from your Drive

  • Agents, to bring their context into the chat

  • Chats, to point at an earlier conversation

  • Knowledge bases, to draw on a whole collection

  • Workspaces and tasks, to ground a request in your wider work

You can reference several things in one message. The assistant sees all of them and can compare or cross-reference across them.

Mention a folder

Reference a folder and the assistant gets its whole tree, not just the top level. If the folder has a SKILL.md, INDEX.md, or README.md at its root, that file is loaded first as a guide to the rest, so the assistant knows how the folder is meant to be used before it reads the contents.

See what you attached

Everything you reference shows up in a preview panel above the message box, each item with an estimate of how much of the model's context it will use. That makes it easy to spot when you have attached more than a model can comfortably hold and to remove something before you send.

Images, audio, and video attach directly when the model supports that type. When it does not, the attachment comes through as a placeholder: the assistant can see that it is there and what it is named, but not its contents. Pick a model with matching capabilities for those files. See Choosing models.

Run a skill with /

Where @pulls up resources, /pulls up skills: reusable instruction bundles you can drop into a message. Type /, choose a skill, and its instructions guide the assistant for that request.

How context works

When you reference a file, the assistant gets a snapshot of its contents at the moment you send the message. If you edit the file later, the copy captured in that message does not change.

Think of it as attaching a copy, not a live link. To give the assistant the latest version, reference the file again in a new message.

Drag and drop

You can drag files from the Drive browser straight into the message box. It has the same effect as an @reference: the file's contents attach to your next message. See Attaching files to chats for every way to bring files into a conversation.

Drag and drop is quickest when you have the Drive browser open next to your chat.

FAQs

Related articles

⚙️

Chat settings and toggles

Customize each conversation from the composer: tool toggles, autonomy, compaction, and a spend limit.

🧠

Choosing models

Browse 200+ AI models, read their cost bands, set the speed dial, and pick the right model for each conversation.

Up next

Web search

AI searches the web during conversations for current, real-time information with sources you can click.

Was this helpful?