Sam Bolgert

gen z engineers will never know the true millennial productivity of blasting emo music while managing 10 coding agents

@ihtesham2005 @cantoni this just happened upon my feed.

Cloudflare just open-sourced an email client where an AI agent reads your inbox, drafts your replies, and never sends anything without your permission. It's called Agentic Inbox. It runs entirely on Cloudflare Workers. Zero third-party servers touching your emails. Here's what the AI agent can actually do: → Reads your inbox and surfaces what matters → Searches across all your conversations → Drafts replies automatically the moment a new email lands → Holds every draft until you explicitly confirm it never sends on its own → 9 built-in email tools running on Workers AI Each mailbox lives in its own Durable Object with a SQLite database. Attachments go to R2. Your email never leaves your Cloudflare account. One click deploys the whole thing. Cloudflare provisions R2, Durable Objects, and Workers AI automatically. Apache 2.0. 100% Opensource github.com/cloudflare/age…

@cantoni @Superhuman has AI features and pretty good management features

Do I know anyone who's using AI to triage their inbox? I'm excited to do this (mostly to get rid of spam) but still nervous about handing over the keys.

@GChase Wasn't familiar with the project until yesterday. Codex got it running in docker with no issue. Hardest part was using a shared cloudflare tunnel to expose it to the internet

@linuxlewis I've tried nextcloud a few times, never got it working how I want (pre claude)

Self hosting is so fucking back. Last 24 hours with Codex on my Debian box: - Samba - Nextcloud - VNC - RDP - new personal site - internal server directory - portable dev environment Tailscale + Cloudflared make self-hosting feel unfairly easy. Codex is goated.

Anyone have a token leaderboard for internal AI use at their company? Doesn't seem possible with either anthropic/openai team accounts

@simonw You gotta put some investment in with an existing codebase to build up your harness and knowledge graph to get consistent outputs. The agent can help with that though so it's easy.

Is there still a widespread belief that LLMs and coding agents are good for greenfield development but don't help for maintaining large existing codebases? I don't think that idea holds up any more

@jesse_vermeulen Talk to one of my other 5 claudes

honest question: what do people do during the 5-10 min while Claude is running?

Code complexity/style still matters in the age of AI coding. When the code is overly complex AI struggles to reason about if effectively and hallucinates more. We've seen success by using skills to help simplify and refactor code as part of coding agent workflows.

@anothercohen @idanbeck

I'm tempted to finally churn off Slack. We're paying ~$6k/year for 40 people and they just quoted me $21k/year for the business version that includes a BAA (and all the shitty AI features). Incredibly overrated software

What's the most impressive thing someone has built with Paperclip? I've tried a few things but it's just burning through tokens without much value

@Austen Orchestrating and coding agents requires the same problem solving as thinking through software solutions. We have to spend less brain power remembering tooling semantics. Same same but different

So far the "AI will completely eliminate the need for engineers" has been 100% dead wrong

@cantoni Gemini has a good free tier for code review that is worth trying out. So far @cognition 's devin is the best I've tried.

I have Codex and Claude both reviewing my PRs. Codex always gives a 👍 with no comments and Claude always breaks down the PR systematically and suggests improvements. It's a good system!

@OpenAI This is dope but I really wanted this pen. Can you hook me up?

Our existing $200 Pro tier still remains our highest usage option. And as a thank you to our existing Pro users on the $200 tier, we’re extending our 2x Codex usage promo (until May 31st) and we’ve reset your Codex rate limits (yes, again).

andybromberg.com/high-effort New mantra: if it’s truly a top priority, it deserves top-priority effort. Obvious. Annoying. Probably true.

How are others managing a company directory of agent skills? I'm trying to avoid adding everyone in the company to our github account but it seems like a git repo is the preferred way of managing this. Maybe I should build the solution this weekend?

@rohanvarma Can you add premium seats with more codex usage? Usage will be too expensive

We just made it frictionless for teams to try Codex! > New $0 Codex only seat for Codex access that is fully usage-based > Annual team seats are dropping from $25 to $20 per month For each Codex only seat you add to a new or existing workspace, we’ll credit your team $100, for your to $500! 💰 Free to get started and only pay for what you use 🤌🏾 openai.com/index/codex-fl…

Are we all not just running coding agents in full-auto dangerously skip permissions mode? When you're running 5+ agents at once it's more cumbersome to juggle the constant approvals. I don't care if you read a file bro

@realfrankwilder What does the content look like?

We replaced our marketing team with an AI agent named Jean Clawd 2 million views in 30 days $0 in ads here's the exact setup 🧵