Claude Tag is the next evolution of agents. It's a proactive, multiplayer agent with memory and identity, built on top of Claude Code. Learn more about how Claude Tag works and best practices for using it in this deep dive.

My first multiplayer experience was filled with team kills. I wonder what Claude's experience will be like?

Again with the filler, just tell us the progress on Fable please

Please bring back Sonnet 4.5 🥺 #sonnet45 #keepsonnet45 #bringbacksonnet45

Wouldn’t it be better with a screenshare? Hard to understand like this.

It’s ironic that I’m currently on the beach trying to use Dispatch that tells me that it can’t help me use my Mac remotely, because to use it remotely via Dispatch I apparently need to unlock my Mac first. That’s all a hoax, none of your remote functionality is working, alas.

FABLE 5 😭😭

for teams using this, what’s the highest-leverage workflow so far? we already have an agent wired in, wondering if it’s worth the time

Fable 5 when?

Can we get that 24 hour fable update?

Hey,update Fable information plz. not these SSAS.

Multiplayer agents break down as soon as two instances prompt conflicting file edits. How does the runtime handle concurrent git writes?

Man - the comments. Anyway…this is something that will help with my clients I am working with and something they have been asking for. It takes a massive step out of the process. Excited to start working with this one!

We're in Brazil's construction sector. Locals rarely use AI tools—it's not industry norm—but all live on WhatsApp. So I built it there: me + 2 women running a R$30M proj entirely via WhatsApp. Key win: onboarding non-tech users to AI w/o training. No DMs. For each, I create a grp: me, them, Iris (my AI). Iris lurks silently till /command. Teach by demo: Paola handles finance & can't prompt. I assign task, then "/Iris, draft doc per Paola's req." Iris delivers; Paola watches. After rounds, she calls Iris solo, gets private thread. All recorded for system learning. Meet ppl where they are: no training, no Slack. Just WA + AI earning team spot.

Things are starting to move in my preferred direction. All the workspace driven agentic applications are winners from this. Having agents in your workspace makes it feel more like you're talking to a real coworker. You're more willing to treat them like people, be patient enough to provide good context, instead of just swearing at them in your terminal.

Claude Tag + Paperclip Desktop for orchestration is the dream stack. Open source Windows app with org charts, budgets, and governance for multi-agent teams.

Genuinely curious what "proactive" looks like in practice does it interrupt you, or does it just mean it can initiate without being called?

Shall we do weekly reset to celebrate?

Can you fix claude? i sent one message and hit 100% usage.

Don't care until full GitLab support.

fable5 is the next evolution of agents, when?

memory + identity is when agents stop feeling disposable

I just renewed my Claude Max 20x plan. Ten minutes ago I was at 1% usage. I opened the app, let Claude read my landing page in read-only mode and instantly jumped to 86% for the entire week. On top of that, the model has been extremely slow since yesterday, with responses taking hours. I just paid and now I feel completely screwed. This isn’t a small glitch anymore. It looks like a serious bug or something worse. I don’t want a support loop. Reset my limit immediately or give me a refund.

This could save a lot of time on routine tracking. The real edge comes from deciding which alerts actually matter.

Bringing AI to a team level sounds great, but the approach is flipped. Team members should have their own AI team working for them, inside their own expert space, removing friction from their work, with access to any relevant business data, on demand (that goes well beyond Slack). Not an AI sitting on top of everything, adding more noise to an already crowded team communication.

the multiplayer angle is what stands out most to me, curious how it handles conflicting instructions when multiple users are tagging the same agent at once, does it queue or does context just get messy

@claude when is fable returning?

Crazy how no-code production coding being seeded through slack. Don’t be surprised if you see No-code Claude tomorrow

Am I tripping, or Claude really come up with something revolutionary?

Stop posting

Curious if there's a resource on how I can configure Claude Tag? If I want to use Thariq's suggestion to create a private channel for CT, for example, how do I configure instructions, memory and tools?

Why not have this with private accounts too? 😬

Fable pls

good to see the goat talk