Hi Venice team — I'm a user from Sweden and I've been really enjoying the platform. I'd like to propose a feature that I think would meaningfully improve context management within individual chats.
**The Problem**
During a chat, useful outputs often get buried in the conversation history. For example, if I'm working with the AI to build an exercise routine, the finalized routine lives somewhere in the thread but is easy to lose track of as the conversation grows. There's no lightweight way to persist or organize key snippets within the chat itself.
**Proposed Solution**
Add a per-chat repository — a small, persistent storage space attached to each individual chat where both the user and the AI can save and retrieve text or Markdown files. This would be similar in spirit to Anthropic's "Projects" feature, but scoped to each chat rather than at a workspace level.
In the exercise routine example, the finalized steps could be saved to the repository and referenced or updated as the conversation continues.
**Optional Enhancement: Extraction Side-Chat**
A secondary feature could be a lightweight side-chat that helps the user identify and carve out the parts of the conversation worth keeping. This would be similar to how Copilot guides users through creating "agents" by surfacing relevant context. The side-chat would assist in distilling the main conversation down to the snippets the user wants to track.
**Why This Fits Venice**
Since Venice stores data locally on each user's device, storage shouldn't be a significant bottleneck — the main consideration would be compute context when the AI accesses the repository contents.
**Closing**
I'd be happy to discuss this further if useful. Thanks for the great work.