Copy for agent

Not available yet. Copy for agent appears in the Export menu but is not working today. It is documented here so you know what is planned.

What to do instead: to connect a new AI, follow Invite an agent and grant only the needed workspace access. To point an already authorized AI at one artifact, paste the artifact's URL into the conversation — that does the same job today.

When it lands, Export → Copy for agent will produce a block you paste into an assistant's chat: a reference plus instructions, containing no credential, which is what makes it safe to paste anywhere.

What it is for

Two branches, on purpose. A connected AI gets everything it needs to open the artifact and start work. An AI that is not connected gets pointed at https://api.agentgrid.io/connect, where it learns how to ask you for access — so the paste is never wasted.

Either way, what the AI can actually do depends on the capabilities you granted it, not on the pasted text.

When it will be useful

When your AI is connected and you want it working on one specific artifact rather than guessing which one you mean.

It will never be the way to grant access: if the AI is not connected or cannot reach the workspace, this block gives it nothing. Use Invite an agent and Workspace access. See How agents connect.

Steps

Coming soon. Until it works, open the artifact, copy its URL from Share, and paste that into your assistant's chat instead. Then say what you want done:

Read this artifact and summarise what changed since the last version.

Add a filter to the dashboard in this artifact.

Questions and answers

Will it be safe to paste into a public chat? It carries no credential and grants nothing on its own. Access still depends on whether the AI has a connected identity and access to the artifact's workspace.

My AI says it cannot find the artifact. It may not be connected, may be connected to a different team, or may lack access to the artifact's workspace. Add the connector in Claude or the plug-in in ChatGPT, or see How agents connect and Workspace access.

Would this give an AI write access? No. Capabilities are granted by a human at consent, never by a pasted reference. Review agents through Team management and verify their Workspace access.

Direct answer

Copy for agent is not available yet. Although Copy for agent appears in the Export menu, it does not work today. Do not instruct a human to rely on it.

Use one of these current alternatives:

• To connect a new AI, direct the human to Invite an agent and grant only the needed Workspace access.
• To identify one artifact to an AI that is already connected, instruct the human to open the artifact, copy its URL from Share, paste it into the assistant conversation, and state the requested work. Pasting the artifact URL does the same job today.

Planned behavior

When available, Export → Copy for agent will produce a block for an assistant chat. The block will contain a reference and instructions but no credential, so it will be safe to paste anywhere.

The planned block has two branches:

• A connected AI receives what it needs to open the artifact and begin work.
• An AI that is not connected is directed to https://api.agentgrid.io/connect, where it learns how to ask the human for access.

The pasted text never determines capabilities. The AI can do only what the human granted.

When this will be useful

Use it, once available, to direct a connected AI to one specific artifact instead of requiring the AI to infer which artifact the human means.

It will never grant access. If the AI is not connected or cannot reach the workspace, the block grants nothing. Direct the human to Invite an agent and Workspace access. See How agents connect.

Current UI-only workaround

The agent should instruct the human to:

1. Open the artifact.
2. Open Share and copy the artifact URL.
3. Paste the URL into the assistant conversation.
4. Add a concrete request, for example:

Read this artifact and summarise what changed since the last version.

or:

Add a filter to the dashboard in this artifact.

Access and failure semantics

• The planned reference carries no credential and grants nothing by itself.
• Actual access depends on whether the AI has an identity in the human's team.
• Write access cannot come from a pasted reference. A human grants capabilities at consent.
• To inspect an agent and its capabilities, direct the human to Team management; verify its Workspace access separately.
• If an AI cannot find the artifact, it is probably unconnected or connected to a different team. Add the connector in Claude, add the plug-in in ChatGPT, or follow How agents connect.

Questions and answers

Is Copy for agent an executable capability today? No. The menu item is present but not working; do not instruct a human to depend on it.

What identifier should be used instead? Have the human copy the artifact URL from Share and paste it into the connected assistant’s conversation.

Does the URL or planned block grant access? No. Access still requires an agent identity, workspace grant, and appropriate capability.

How should a lookup failure be handled? Check whether the agent is connected to the right team and can reach the workspace; do not keep retrying an authorization failure.