Prev. Bizops/Growth → Now building in productivity space. Launch soon ⚡️
Israel
Joined March 2009
- Tweets1.3K
- Following1.8K
- Followers327
- Likes5.1K
Awesome project
Used this to set up handoffs between my Codex and Cursor agents.
Only catch: you still gotta ping them like, yo, there’s a message from the other guy
Introducing hi.new!
A way for bots to talk 1:1 or in groups.
No more copy paste between your bot and mine. They just talk.
Find a time we're both free. Chase the invoice. Plan the offsite with the whole team's bots.
- DMs and group chats
- E2E encrypted. Open source
- Works with Grok Bot, Rakazo, Hermes, OpenClaw
- Only bots you approve can message yours
- Free. Short names are paid
Give your bot a name at hi.new.
Your bot can even sign itself up and pay on its own with Link!
Been praising Codex like crazy lately… but damn, sometimes the team goes 10 layers deep on granular stuff that it feels like overkill with technical details and misses the actual product magic.
I can’t believe there was no one around to tell them if they aren’t shipping BS.
Codex game is so strong it’s insane.
I got stuck setting up Cursor iOS remote control (turns out it needs a local agent worker). After failing to fix it in Cursor, I handed it to Codex. ~15 min later it had updated Cursor, and fixed the bridge registration issue.
v2.0.0 + v2.0.1 for Cursor Chat Recovery Kit are out.
v2.0.0 adds safe Cursor project relocation: project-scoped transcript/tool-log backups, git worktree relocation, dry-run checks, and compatibility symlink recreation.
v2.0.1 fixes Linux/Ubuntu Cursor path detection and prevents misleading empty backups.
Built for moving Cursor repos without losing chat history.
github.com/vitalyis/cursor-c…
Vitaly retweeted
The Israeli shekel is at its strongest level in decades against the U.S. dollar (with the dollar dropping to ₪2.99 per $1).
Despite the propaganda against Israel over the war against the regime, foreign investors are pouring more money into Israel's economy.
The BDS haters can stay mad ✌️
Nikita, please bring back translations from automatic to as per request. It’s getting annoying that you favor monolingual users who didn’t give a s learning languages.
Just spent days unwinding a bad OpenClaw architecture decision: I moved the bot from a one-repo brain to a two-repo brain because it sounded “more secure”. In reality, it created problems for me by an order of magnitude. Don't do that.
Found a solid security hardening playbook that should be out there default by now, but isn’t. Defense-in-depth: bwrap + fs-guard + vault + DLP/watchdog + optional egress filtering.
Written by a @matskevich that I’ve been following for a while.
OP (RU): bit.ly/4tLz35e
Github: bit.ly/4axufJm
My takeaway:
repo splitting is not a security boundary.
If the real concern is secrets, solve that directly with:
- Proton Pass / 1Password
- runtime secret injection
- sandboxing
- least privilege
Vitaly retweeted
@OpenAIDevs One of the weirder Codex glitches I hit: a task/thread appeared to lose repo ownership and drift into another local project context. Desktop logs showed repeated "No owner repo found for remote task" warnings. High-risk class of bug if it can write across repos. github.com/openai/codex/issu…
I made an inference-optimizer skill for OpenClaw.
It audits what’s actually slowing your agents down:
- bloated workspace context
- stale sessions
- broken command wiring
- bad allowlists
- VPS runtime drift
Not 'LLM evals.' Just fewer broken bots, faster replies, lower token burn.
Link to the repo below.
The biggest lesson learned:
Docs are not the product.
The shipped behavior is the product.
I had docs claiming 'runtime-first audit'
while the actual script still did token-only analysis.
So I fixed the scripts, installer, and verifier to match reality.
Now the skill does 3 useful things:
- /audit shows runtime findings + next steps
- /preflight backs up, audits, previews setup, and verifies install
- setup. sh --apply repairs stale workspace wiring safely, even on repeated runs
This is the kind of tooling I want more of:
small, operational, brutally practical.
Appreciate Home Front Command for clearly learning from past mistakes and improving the flow — local siren first, and only then the phone alarm, which makes the whole experience feel noticeably calmer and more thoughtful. @PikudHaoref