Using Claude with Hive: what it can touch, and as whom
A few of you connect Claude to Hive, so here is how the connection works now.
You sign in with Hive when Claude asks. No API key goes in a URL any more. Claude then works as you, with the permissions you already have in your workspace and nothing beyond them. It can read your projects, create actions and follow-ups, update statuses and draft comments.
If you connected before the end of July, remove the old connector and connect again. The setup steps are at https://hive.com/integrations/claude
If it got in your way, I'd like to hear about that as well. What have you used it for so far?
If you'd rather set it up by hand, the setup page also has a manual route: in Claude's Settings, under Connectors, add a custom connector with the Hive MCP URL given on that page. https://hive.com/integrations/claude
I'd start with the directory route, because it has fewer steps: open Hive in the Claude directory, click Connect, then sign in with Hive and approve access.