Wendy Zhao

Hi Tomas, I understand your frustration with the Confluence folder sync not being available. The main blocker is the technical limitations with Confluence's API: Confluence has very aggressive rate limits (some endpoints are limited to 25 requests per 15 minutes), and supporting folders would likely "10x worsen the situation" according to the engineering team, while Confluence released folders as a beta feature in September 2024, their API is still incomplete - meaning that you can only retrieve a folder by its ID, but there's no way to list all folders. The current API lacks proper batch support, requiring individual API calls for each page/folder, which exacerbates the rate-limiting problem. The team has been actively working on this with Atlassian's Confluence team With that being said, unfortunately, there's no specific timeline for when this will be available.

here is where you can find the full context: https://dust-tt.notion.site/Slack-Terms-of-Service-update-and-API-Changes-21728599d94180f3b2b4e892e6d20af6

Sorry for the confusion, there are new Slack tools, but they work very differently from the previous data source approach. allow me to clarify: what is in the past: Slack as a data source: The ability to have Slack channels indexed and searchable through semantic search in your AI agents after the change: the agents can no longer search across all your Slack history using AI-powered semantic understanding What the new Slack MCP tools provide:

the key differences being: Instead of having all Slack content indexed and immediately searchable, agents now make live API calls to search Slack, and the search results are limited to what each individual user can see, not workspace-wide access Dust is working with Slack closely and will keep the community informed 🙏

Hi Raphael, sorry for the delayed response (I am not actively on the community Slack channel, really apologize about that, feel free to drop us an email at support@dust.tt for this sort of issue) , we have recently changed the max steps taken to agent for 128 steps by default for all now, i will look into this today 🙏

Hi, happy to clarify - Yes, the Slack terms update significantly impacts Slack channel data sources in AI agents. Here's what's happening:

Impact on AI agents:

As of July 25th: Current @Dust bot renamed to @DustLegacy , on August 1st: @DustLegacy stops responding, and on September 2nd: Slack DataSource completely deleted , we will communicate before this happens In short, you should migrate to the new Slack bot app from the marketplace for @Dust interactions, but you will lose the ability to use Slack channels as indexed data sources for your AI agents after September 2nd.

Hi there, this error message shows that the connection to CoreAPI was interrupted, you can try switching to a different AI model for your agent diligenceAnalyzer and see if it helps?

Hi, sorry for the delayed response - our team identified and fixed the issue, could you please try gaain?

Hi all, the Dust Chrome extension does not work with Arc browser because Arc browser doesn't support Chrome's SidePanel API, which is essential for the Dust extension to function. The extension relies on this API to display the Dust interface as a side panel in your browser. Unfortunately, Arc has explicitly stated they won't be adding support for the SidePanel API in the near future. When users have contacted Arc support about this, they've received responses confirming they "aren't going to be adding support for that API in the near future." 😕 You could consider using dust.tt directly or using regular Chrome/Chromium browsers, where the extension works perfectly

Hi there, confluence folders aren't supported for sync yet due to confluence's api restriction . We'd recommend you to move the important files out of the folder for the time being, the team is aware 🙏

Hi sorry for the delayed reply, it was a piece of generic recommendation 😅 I hope the rest of the tips can still be helpful though