In Karatsu, Japan, for an official Anthropic Claude Community hackathon. Built and deployed Sea Forest Report in under 45 minutes, including planning and full code reviews.
And the venue? PLAPLA is something else. Hard to beat building by the sea.
@thsottiaux Would love to help. Meta certainly has removed all friction from sharing Muse. Steal that. Not sure why you call the product Dots when the experience is clearly Dot. There’s some minor stuff that you can learn from Muse and Grokbot if you want to run up to 1B users. Hit me.
Results: same 5 PR plan to Cursor 2.5 and Meta Muse 1.3.
Cursor, 5:45;
Muse, 11:30.
Nearly identical code. Cursor, duplicate PRs. Muse rewrote backlog to cover tracks. Picked 4 Cursor, 1 Muse. Coding is fast as fuck now. Getting agents to follow instructions? Not so much.
@harden_369741 Don’t need to skip it. Make it part of the initial prompt/plan. If you’re coding with Cursor or Claude you can do a review with Agy or Codex.
And if you really want to automate, I’ve found (meta) muse agent to be useful with doing read-only code reviews. Least destructive 😂
Anthropic released code reviews, which is great since Google sunset reviews in July and Codex reviews on GitHub are just terribly slow. Would love to try it but apparently it's only available to enterprise customers so I guess Claude... I understand we aren’t devs in your eyes.
@WheelerMan This is a major outage now holy how many hours has it been? I don’t know how many people feel screwed over but it looks like I’m going back to Hermes ASAP.
@elonmusk Almost got arrested for bringing a sat phone to India. They’re mostly afraid of the people organizing and taking back power. Or what they like to call a “terrorist threat.” You’re barking up the wrong tree.
Ask an AI coding agent to move a button and somehow you’re reviewing a pitch deck for a company you didn’t start.
New framework. New logo. Three pricing tiers.
The button hasn’t moved, but it’s raising seed funding.
Next Wednesday, bring your agent and a real sales or marketing task to AI Braintrust’s Agents for GTM.
Outreach, content, lead follow-up, partner research. Bring something you actually need done.
Apparently in a completely idiotic move, @AnthropicAI has banned cruelty towards Claude. In the 2026 usage policy update dated October 8, 2026, there is a section called "Addressing Abusive Behavior Towards Our Models," which is hilarious and completely out of touch. #SMH
Looks like #dots got announced early. RIP #OpenClaw. So now you have GrokBot, Muse, Dots and Hermes to contend with for the serious players in the Agentic space.
https://t.co/XlPYrkoGt3
Yes, it looks like #Claude is down for a good proportion of users, but the real OG move is whether or not you got a high score. The most I ever got was x6, but I was only able to screenshot and reproduce an x4. What's your high score? Looks like this is linked with their new ToS.
@MarcNixon24 No quite normal. Parliamentarians in Europe just don't show up. It's complete grift, and that is why it's so concerning for Canada to align with the EU. Just so much waste and disconnect.