Pinned Tweet
Claude Code just told me my input and output since April add up to about 1,908 copies of Anna Karenina.
179.5 billion tokens. 2,700 sessions. 184 of the last 221 days.
Most of that wasn't me typing. It was Claude reading the same codebases again and again while we built products for clients, my own SaaS and training material for teams across Kenya.
The number isn't the point. The point is that one person in Nairobi can now ship like a small team, if they show up every day.
What would you build with that much help?
6
2
95
We tried something different with Claude Community Kenya on Sunday. No laptops, just a walk through Karura Forest and a lot of conversations about AI.
1
23
Back to building this week: Claude CAD Workshop on Thursday and the Claude Impact Lab on Saturday, both in Nairobi. Links below.
1
8
Register here:
Claude CAD Workshop, Thu 19 Nov: luma.com/claude-t9v5
Claude Impact Lab, Sat 21 Nov: luma.com/claude-ptdd
6
Tomorrow in Nairobi we're taking the Claude conversation off the screen.
A walk through Karura Forest, small groups, swapping how we actually use Claude. Sun 2 to 5 PM, Gate A.
Limited spots, register via the QR on the poster.
What would you ask another builder on that walk?
22
Nairobi, tomorrow 2 to 5 PM: Claude Conversation at Karura Forest, Gate A. Walk and talk AI with people building with Claude, then sit and keep going. Entry on us, plus a cap, notebook and pen. 30 spots. luma.com/claude-sznz
1
1
36
Nairobi, tomorrow 2 to 5 PM: Claude Conversation at Karura Forest, Gate A. Walk and talk AI with people building with Claude, then sit and keep going. Entry on us, plus a cap, notebook and pen. 30 spots. luma.com/claude-sznz
23
Anthropic is coming for the open source agent stack. Not by fighting it. By taking away its best argument.
Most teams self-host agent frameworks for one reason: "our data can't leave the building."
Since May, Claude Managed Agents answers that:
🕷Tool execution runs in a sandbox you control (your infra, or Cloudflare, Daytona, Modal, Vercel)
🕷Files and repos stay inside your perimeter
🕷MCP tunnels connect agents to private internal systems without exposing them
Anthropic runs the agent loop. You own where the work happens. That split is the whole play.
Honest caveat: sandboxes are public beta and tunnels are research preview. Not production-ready for everyone yet.
But think of a Kenyan bank or insurer where data residency kills every AI pilot. That objection just got a lot weaker.
The future of AI agents isn't just better models. It's who controls the environment they run in.
#BuildersNotConsumers #ClaudeCommunityKenya
Self-hosted Claude Managed Agents is now in private beta.
You can host the agent harness, sandboxes, and session history, in your own cloud or on-premises, and choose where requests to Claude are sent.
78
Introducing Claude Meetings: add Claude to your team’s video call and it builds docs, whiteboards, and other artifacts with you while you’re still talking.
Starting to roll out in beta on Team and Enterprise plans.
Spidey(Peter Kibet) retweeted
We've given all Pro, Max, Ultra, and Team users one free full reset, across all models and limits. Use it anytime before December 10 at 2:00PM PT.
We recently fixed a few small issues that used more limits than intended, and the Fable resets didn't do much for those of you who'd already hit your weekly limits for all models. While these issues didn't affect everyone, we want to make up for the trouble.
You'll find it in Settings > Usage, or at clau.de/reset.
Read more about how Claude designed these devices: anthropic.com/research/claud…
During pre-release testing, Claude Fable 5.2 and Claude Mythos 5.2 worked on ambitious projects in many different fields. Here are a few.
From a one-paragraph prompt, Fable 5.2 designed three working electronic devices, including this home voice assistant hub.
Spidey(Peter Kibet) retweeted
The future is now.
Spidey(Peter Kibet) retweeted
If you're someone like me who developed a reading habit recently, you might notice that some books sound like AI. But these books existed long before AI became dominant, so really, AI sounds like books.
It seem like the em dashes, two-part titles, antithesis have been used for years. If this is how people have always crafted sentences and paragraphs, why do we suddenly have a problem with them?
Tracy built a flood watch in one Saturday. Mohamed built a fleet manager for Mombasa's tuk-tuk owners.
That was our Claude Build Day at @wsh_mombasa on 31 Oct. Real problems from the street.
In Mombasa? Join us for the next one: chat.whatsapp.com/HinlY1qyzY…
55
Spidey(Peter Kibet) retweeted
October was a big month for Claude Code 🍁
We shipped mods, made the desktop app better at running lots of sessions side by side, gave you more insight into where your usage goes, and plenty more. Here's a recap!

















