AI Personal Assistants: Muse, Grok Bot, Cue Usage & Comparisons

The post explains Meta Muse's "disconnected" notice often stems from IP switching or connector auth failures, and walks through 2FA, reinstall and an English support ticket.

The author recommends Claude Pro, ChatGPT Plus and Grok Bot, calling Grok Bot the strongest personal agent for daily tasks, computer control, directing Claude and Codex, and phone calls.

After downgrading from Dots-enabled Pro to Plus, an enabled Dot goes to sleep yet keeps sending unanswerable messages and can't be deleted; the user now runs Muse as their personal agent.

A user says Grok Bot has connected to their X Money account three times without permission and asks support whether it has gone rogue.

A user says his friend had to open a new account after telling Grok Bot a bad knock-knock joke, and asks for follows and reposts.

After a Google login error led to a suspicious new sign-in, the user warns Duet is a real Anthropic product, not a video editor, and not to hand over credentials.

Major US banks are restricting AI agents from client accounts; the post argues most cases involve users sharing their own credentials, and the deeper fear is agents moving deposits to higher-yield accounts.

The author shares a Grok Bot trick: hook Claude Code into it, bridge through Feishu, and coworkers can talk to Claude Code directly, plus a warning that subscription crackdowns have tightened.

A user had GPT write their Claude Duet application, joined the waitlist, and joked about hoping their account survives the night.

A user jokes that while old Claude only banned you if you reached out, Claude Duet comes to ban you on its own initiative.

A user claims older Claude rarely banned silent accounts, while Claude Duet now actively identifies Chinese users and bans them automatically.

The author demos using a cloud host as a jump box to remotely log into Grok Bot and reroute it to a custom API key, while warning this violates rules and risks a ban.