Anurag Yadav (@codeStunners) header
Anurag Yadav (@codeStunners) profile picture

Founding Engineer @ Celesto AI . Building Infra for Agent Era

Bangalore India
Joined May 2023
Photo by Anurag Yadav (@codeStunners)Photo by Anurag Yadav (@codeStunners)Photo by Anurag Yadav (@codeStunners)Photo by Anurag Yadav (@codeStunners)Photo by Anurag Yadav (@codeStunners)Photo by Anurag Yadav (@codeStunners)Photo by Anurag Yadav (@codeStunners)Photo by Anurag Yadav (@codeStunners)Photo by Anurag Yadav (@codeStunners)Photo by Anurag Yadav (@codeStunners)
Anurag Yadav retweeted
Playwright and browser security reminder. We saw a post suggesting that you delete Playwright because it bundles browsers with CVEs. That is a misconception. Do not delete Playwright🙃. Here is how it works and what you should do. No Playwright package bundles a browser. Agents on Playwright MCP or CLI run your stable Chrome, which gets Chrome's security updates, with the sandbox on. 🤖 If you use Playwright MCP or the Playwright CLI They use the stable Chrome / Edge on your machine, so it gets its normal updates. No config edit is needed. 🧪 If you use the test runner When you run npx playwright install, Playwright downloads Chrome for Testing, pinned to the build we tested that release against. The pin is on purpose. Your tests run on the same browser on every laptop and CI machine, and a browser update cannot suddenly break your suite. You choose when to move by upgrading Playwright. New releases usually pin a build ahead of stable Chrome, so you see breakage before your users do. We recommend this for full e2e test suites on your own trusted app. This build can miss Chrome security fixes until the next Playwright release, and by default it runs with the sandbox off. Do not use it on the hostile web. 📚 If you use the library API chromium.launch() has the same defaults: the pinned build with the sandbox off. If your code visits pages you don't control (scrapers, agent scripts), max out protection: chromium.launch({ channel: 'chrome', chromiumSandbox: true }) 🔒 Always keep Chrome / Edge updated. Some bugs can escape a sandbox. Happy testing!
44
276
33
2,354
170,324
Anurag Yadav retweeted
Replying to @izadoesdev
Shifted to @CelestoAI and been great so far
2
3
198
Replying to @owengretzinger
Switch to @CelestoAI Life’s good and affordable
1
1
2
4
294
Anurag Yadav retweeted
20,000 CI minutes a month on a 2 vCPU / 8 GB runner: GitHub: $120 Blacksmith: $80 Ubicloud: $40 Celesto CI: $13.33
15
3
3
97
18,012
Anurag Yadav retweeted
Your coding agents open PRs all day. Every one of them kicks off CI. Celesto CI runs your GitHub Actions on fresh microVMs from $0.00025/min, nearly 90% cheaper per minute than GitHub-hosted runners. Install the GitHub App and get $5 in free credits.

2
8
50
463,204
Anurag Yadav retweeted
manifesting YC W27
9
1
58
1,288
Anurag Yadav retweeted
Today was my last day at @Railway! I was let go after introducing a major security issue triggered due to my agents went off the rails. From today, I'm gonna use Celesto sandboxes for all my agents.
5
2
1
43
6,232
Anurag Yadav retweeted
Celesto is 6x cheaper than Blacksmith
Replying to @aniketmaurya
What about vs blacksmith?
3
1
21
781
Anurag Yadav retweeted
Two critical security vulnerabilities affecting PHP have been identified that result in unauthenticated remote code execution (RCE) for any PHP application. Ubuntu has released fixes for the affected packages for all supported releases. Read more: ubuntu.com/blog/php2shell-vu… #Ubuntu #Security #CVE
32
398
33
1,653
66,663
‼️BREAKING: PHP has released its fix for the two leaked flaws, which can let attackers run code on servers without logging in. "There are no complete mitigations, updating is the only safe choice," PHP says. Update to PHP 8.5.12, 8.4.27, 8.3.36 or 8.2.35 now. PHP says Apache with mod_php is most at risk, with a demonstrated exploit that fires before any application code runs. PHP 8.1 and older get no official update, though PHP has published source backports for 7.4, 8.0 and 8.1.
63
968
113
4,587
257,663
Anurag Yadav retweeted
🚨 LiteLLM, a Python package for connecting to AI model providers with 92M+ monthly downloads, has been compromised. Version 1.107.0 ships credential-stealing malware through PyPI wheels and Docker images, targeting cloud credentials, developer secrets, and AI tools.
110
324
178
1,906
304,873
Anurag Yadav retweeted
Anthropic pushed 25 million code checkouts per day to realize that they need to rebuild "git". A technology that every company depends on for building their software. Computer infrastructure built for humans now fail for agents and agent assisted work. Coding agents has completely changed the way I used to work. I've 5-6 agents working on independent tasks while I research a complex task (also with AI assistance). A Rust compilations took 15 minutes per code checkout. Around a thousand of code commits would have costed $90, Celesto reduced it to only $4. Furthermore, it reduced the total compilation time as well -- faster processing as compared to GitHub. Great bang for the buck if you say!!! I've built a cost calculator for your company as well - celesto.ai/ci
1
1
1
17
1,194
Anurag Yadav retweeted
Hell yeah!! Hosting first hackathon in SF
2
1
34
790
Anurag Yadav retweeted
A customer who signed up yesterday has already ran 496 minutes of CI with Celesto 🤩
1
13
519
Anurag Yadav retweeted
Users created more Celesto sandboxes in the past week than in the entire previous month.
2
1
20
1,100
Anurag Yadav retweeted
We ran the same CI job on GitHub Actions and Celesto. Celesto finished 23% faster with half the CPUs. ⚡ Here's the benchmark (cache hit): - GitHub: 31s (2 vCPUs) - Celesto: 24s (1 vCPU) We just shipped dependency caching for Celesto CI. ✅ Works with existing GitHub Actions workflows ✅ Supports `actions/cache`, npm, and pip ✅ No separate cache charges Your workflows stay on GitHub. Your CI runs on Celesto. Read the benchmark and how we built it: celesto.ai/blog/posts/celest…
1
1
9
532
Anurag Yadav retweeted
I built a cost simulator for GitHub Actions checkout here - celesto.ai/ci
1
1
7
312
Anurag Yadav retweeted
just ran @computesdk benchmarks on Celesto. we really are coming for everyone
2
1
1
11
705
Anurag Yadav retweeted
Every AI agent has its own computer. What a mess. Claude Code has one. Codex has another. Hermes has another. 3 fckn computers to manage. And I have tens of bots running. Why can't they just share ONE computer? Well, now they can. ⚡ We just shipped Celesto MCP. Claude Code, Codex, Hermes, and other AI agents can now use the same cloud computer. Switch agents, not computers.
29
1
4
25
1,357