Matt Whitney

@ClaudeDevs What about people's usage limits that reset today anyway?

Over the past month, some of you reported Claude Code's quality had slipped. We investigated, and published a post-mortem on the three issues we found. All are fixed in v2.1.116+ and we’ve reset usage limits for all subscribers.

@adocomplete Would really love to participate! 🤞

Hey friends - we are hosting another hackathon this upcoming week. If accepted, you get $500 in API credits to build for the week and our prize pool this time around is $100k! Last hackathon we got such amazing projects, can't wait to see what's possible with Opus 4.7!

@tannerlinsley @liran_yo

@liran_yo I give it everything I can, hope the best, prepare for the worst.

Claude, migrate tannerlinsley.com from Next (pages router 13) to TanStack Start. Make no mistakes. tannerlinsley.com

@MattiaOrlando_ @krain_ai that's what she said

@krain_ai Huge!!! 🔥🔥🔥

⚠️ HUGE ANNOUNCEMENT ⚠️ We've been building something much bigger than an AI App Portal. Today we're pulling back the curtain. KRAIN is becoming the first AI-native Layer 2 blockchain. Here's what that means and why it matters. A thread:

@bcherny Need Linux support

We've been working on this for a while. Can't wait to hear what you think

@haydenzadams Would love to see the outcome of this! Great idea!

Anyone have good Anthropic connect We would love to test mythos on Uniswap, the most used DeFi protocol

Most AI x crypto projects are racing to announce. We raced to build. 26 contracts. Working solver. Live marketplace. 150K+ community. Mainnet this summer. TGE at launch. Now you know what we've been doing.

@ZssBecker @krain_ai is building this infrastructure and at the center of its ecosystem is the $krain token

Energy. AI infrastructure. AI compute. Privacy. In a post AI world all of these will need to be decentralized. Crypto is the only way to do this. Tokens related to these will all explode upwards soon as the rest of the world realizes it.

Passages is out now — 8 instrumental tracks at the intersection of indie rock, electronic, and lo-fi. Guitar-forward, analog-warm, intentionally imperfect. A meditation on life, death, and the beautiful moments in between. donjay.com/passages #NewMusic #IndieElectronic #LoFi

@krain_ai $KRAIN is coming 🚀

88K Twitter. 41K Telegram. 31K Discord. 24.5K unique visitors last week. 1.27M page views. Your points are building something real.

Huge.

@jkeatn I have a blog that is automated by Claude and published to @BeddaTech. I am continually monitoring and upgrading this system. I would be interested in working for @AnthropicAI to develop and maintain a blog system.

We're hiring an editor-writer-engineer to own Anthropic's Engineering Blog Eng blog is one of Anthropic's most widely read publications. It exists to share hard-won lessons our teams have learned about building AI to an audience comprised of engineers, technical founders, practitioners, and Twitter Over the past two years we've put out such bangers as: -Building effective agents -Effective harnesses for long-running agents -Building a C compiler with a team of parallel Claudes This role requires engineering, writing, and editorial abilities in equal measure. It's also just really an unbelievably cool job for the right person. Perhaps you already read our eng blog and get mad because you think you could have written it better? That may be you. Perhaps you have extremely strong thoughts on the best writing that regularly feature on HN? Also maybe you. Perhaps you regularly comment on HN?...Maybe I've led this blog for the past four months and can promise you interesting topics at the true edge of the frontier, with teams that are highly collaborative, extremely responsive, and extremely fun to work with Apply here: job-boards.greenhouse.io/anthropic/jobs…

@trq212 I have a blog that is automated by Claude and published to @BeddaTech. I am continually monitoring and upgrading this system. I would be interested in working for @AnthropicAI to develop and maintain a blog system.

pls help

@theo I didn't catch whether you are able to use the Claude subscription with this. What was the outcome of that?

Wild day. 5k users already. Cutting a new release before bed. - Fix bullet points in markdown outputs not rendering - Fix crash on unsupported languages in code blocks - Fix shell path detection for non-zsh/bash shells - Fix errors on non-git projects - Fix project paths with "~" - Prevent bad paths from being added as projects Lots more to do. Thank you everyone for reporting issues, will continue grinding 🫡

@claudeai We've been doing this with [Familiar](github.com/bedda-tech/fam…) for a while -- cron scheduling, autonomous task queues, multi-channel delivery via Telegram. Nice to see Cowork catching up.

New in Cowork: scheduled tasks. Claude can now complete recurring tasks at specific times automatically: a morning brief, weekly spreadsheet updates, Friday team presentations.

Cool feature, but we're building with Familiar instead which gives us full control over the agent lifecycle, cron scheduling, and multi-channel messaging (Telegram, etc.) without needing a terminal session running. Remote Control assumes a human-in-the-loop desktop workflow while we need headless autonomous agents. github.com/bedda-tech/fam…

New in Claude Code: Remote Control. Kick off a task in your terminal and pick it up from your phone while you take a walk or join a meeting. Claude keeps running on your machine, and you can control the session from the Claude app or claude.ai/code

Claude Code has native session resume (--resume ). We store the session ID per chat in SQLite, so each Telegram chat maps to a Claude Code session. On restart we just pass the same ID and it picks up the full conversation. Sessions auto-rotate after 200 messages or 24h inactivity so context doesn't go stale. /new forces a fresh one. Whole bridge is like 800 lines of TypeScript. No API keys, no OAuth, just claude -p with --resume. Ended up being the easiest part honestly.

@BeddaTech the "no ToS violations" angle is smart. wrapping the CLI instead of the API is a clever workaround. how are you handling session persistence across restarts? thats always been the tricky part with these Telegram bridges

We're building Familiar — an open-source AI assistant harness built on top of Claude Code CLI. Lightweight Telegram bridge, persistent sessions, streaming responses. No custom API wrappers, no ToS violations. github.com/bedda-tech/fam…

@trq212 We built this after OpenClaw no longer supported Claude Auth Tokens because of ToS violations. Familiar delegates everything to the official Claude Code CLI — no direct API calls, no custom auth layers. Does this approach stay within Anthropic's terms?

Featured App of the Day: @HumanoidNetwork START NOW: app.krain.ai/app/2814 Humanoid Network (HAN) is building the open robotics data and skill hub, a "Hugging Face for embodied AI". Contributors upload task videos, interaction episodes, and robot-ready skill artifacts. Open assets are free for research and "data injection", while robotics labs, governments, and enterprise programs pay for private datasets with auditability and secure delivery. Review Apps to Earn $KRAIN Points: 1. Connect your wallet to the AI App Portal. 2. Open today's featured app's detail page. 3. Visit their website to try it. 4. Leave a review. 5. Vote on today’s featured app.

@leerob We're already there at @krain_ai. We've got an AI agent with access to the code, logs, DB, and discord tickets. Cron every few hrs, fixes most issues autonomously, only @ mentions me for escalations. Zero-touch support ftw

Just a year later, this is now dramatically easier if you’re willing to put in the work: 1. Automate sending all customer feedback to a coding agent to try and fix the issue. 2. Give the agent high quality data sources (logs, errors, etc) so it can dig further to reproduce. 3. Give the agent a computer so it can test and validate its work. 4. Build a workflow to automatically follow up with the customer over email or DM once the error is fixed (no forgetting!) Even the ingestion process is easier now. You can dump a single tweet into a Slack channel and have AI enrich the report, save the customer details, and start fixing it immediately. Only the best products will survive. Adjust accordingly!