FAQ

It broke. Or it looks like it did.

Thirty-one questions, grouped by where you were when you asked. Every answer starts with the short version in bold; open the card for what to do and why it matters. Most of the things that look like bugs in Berd are the product working as designed — the answer says which.

• Getting started

  ---

  • Which providers and harnesses can I use?
  • I saved my key and got an error — what does it mean?
  • Do I need Goose?
  • Does Berd run on Windows or Linux?
  • I skipped the tour — how do I get it back?
  • What is the usage-data toggle at welcome?

• Agents and skills

  ---

  • Where are my agents and skills stored?
  • My own skill replaced a built-in one — is that a bug?
  • Agent, skill, project, automation — which is which?
  • Can I make my own Gloopie?
  • Why does the Discover tab ask for a CLI?
  • Can I give an agent to a colleague?
  • Which built-in agent should I start with?

• Working in sessions

  ---

  • I sent messages while the chat was starting — are they lost?
  • What does Steer do, and why is it unavailable?
  • What happens when the conversation is compacted?
  • Where did my files and changes go?
  • Can I switch model mid-chat?
  • Can I search old chats?
  • An app in the chat wants to send a message as me — should I let it?

• Automations, CLI and advanced

  ---

  • My automation failed — is it the provider or the automation?
  • Can my agent change Berd for me?
  • What is the difference between feedback open and feedback submit?
  • Should I turn experiments on?
  • What does the monitor skill do for me?

• Privacy, safety and Buzz

  ---

  • Does Berd send my data anywhere?
  • Can Berd read my Buzz channels?
  • Should this work go in Berd or Buzz?
  • Why can't I edit a company-managed connection?
  • How do I move Berd to a new laptop?
  • Can I contribute a fix?