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

# 0.3.10

> Run Claude Code on Cogno credits, choose meeting viewers, and get more reliable execution.

## TL;DR

## Claude Code on Cogno credits

Claude Code can now run on Cogno credits alongside Codex. Select Claude Code with the **Optimized model** in the conversation’s agent picker when the Workspace has credits and the connected Engine supports it.

Background work can also use **Optimized** model selection. The completed turn shows the selected model, and the usage log identifies the agent used for each credits run. See [Agents](https://cogno.studio/docs/agents.md).

## Meeting viewers

Choose who can view a meeting and save that selection independently of Project assignment. Open the meeting’s row menu to change access settings. See [Meetings](https://cogno.studio/docs/meetings.md).

## Reliability improvements

Cloud work can continue across Projects. Completed work reports its result while file saving retries separately. Execution and connection states show when confirmation is still pending. Unanswered requests and completed runs without a submitted result are now visible.

Questions reach the Task’s assignee. Login can be retried after an OAuth failure, conversations stay at the latest message when resized, and Desktop can recover after its web content process stops. Desktop also checks for updates while it remains open.

## Related docs

* [All releases](https://cogno.studio/docs/release-notes.md)
* [Agents](https://cogno.studio/docs/agents.md)
* [Meetings](https://cogno.studio/docs/meetings.md)
* [Cloud environments](https://cogno.studio/docs/cloud-environments.md)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.