Ask the assistant for a podcast on any topic, and it handles the entire production. For news and current events it researches the recent international coverage; for timeless subjects it gathers general material. Then it writes a three-part narrative, turns it into a script of about thirty minutes, and records the narration with its own voice engine. The result is a ready-to-publish MP3 file, complete with an opening intro, background music, a host farewell, and a closing sound, plus an RSS feed that podcast applications understand.
You do not need to prepare anything. Describe the subject and the tone you want, and the assistant does the research, the writing, and the recording. The episode appears in the podcast folder of the program, and the RSS feed makes it easy to distribute to listeners who use podcast apps.
The narration uses the same voice engine as the chat and the phone, so the podcast sounds like your assistant. Everything is produced on your machine: the research, the text, and the recording never depend on an outside service.
The next guide explains the protections built into AgentBridge: the sandbox that keeps the assistant inside its allowed area, and the privacy measures that keep your data yours.