@mikeinweb Testnets are removing friction from deployment, not from responsibility. The gap between shipping and understanding is where most failures will come from.
๐ฅWARNING: Vibe coding without a Lead Architect prompt is throwing away 80% of your results.
How do you vibe code like a senior engineer?
Here's what a structured AI prompt actually gives you:
โ Enforced separation of concerns across every file it writes
โ Named directories, stated before a single line of code
โ Auto-typed functions, docstrings, and error handling every time
โ Tests suggested after every implementation
โ Security rules baked in: no hardcoded secrets, rate limiting, input sanitization
Without this? AI Coding agents writes code. With this? they thinks like your senior engineer.
AI coding tools fail at architecture. One prompt that enforces your structure turns a vibe coder into a disciplined staff engineer.
You do not need a Mac Mini to run an AI agent.
People are running autonomous agents on:
โ $5/mo VPS servers
โ Old laptops
โ Cloud instances
โ Raspberry Pis
The barrier to entry is lower than you think.
The hard part is not hardware. It is designing what your agent should actually do.
AI canโt yetย one-shot an entire productโbut with the rise of vibe coding, supercharged developers are getting close.
Part intuition and experimentation, part validation, it's the new frontier of software development.
YC's Tom Blomfield (@t_blom) has spent the last month building side projects with tools like Claude Code, Windsurf, and Aqua, seeing just how far you can push modern LLMs. From writing full-stack apps to debugging with a single copy paste of an error message, AI is becoming a legit collaborator in the dev process.
With tips from YC founders in the Spring batch, this is a playbook for anyone who wants to get the most out of vibe coding and build faster.
๐๐ก๐๐ฆ๐๐ฅ๐๐จ๐ง ๐๐๐ฏ ๐๐ฉ๐๐๐ญ๐ โ ๐๐๐ซ๐๐ก ๐๐ โ ๐๐ฉ๐ซ๐ข๐ฅ ๐, ๐๐๐๐โฃ
This week the core team ran the internal testnet.โฃ
โฃ
๐๐ง๐ญ๐๐ซ๐ง๐๐ฅ ๐๐๐ฌ๐ญ๐ง๐๐ญ ๐๐ฎ๐งโฃ
Deployed 3โ5 validator nodes using the new testnet chain spec and ran the install script end to end. Validated block production, staking, emissions, pDEX swaps, and bridge deposits. Also validated the full privacy transfer flow, shield to confidential transfer to unshield, all on real testnet parameters with the 100M CHML supply and 1,750 pCHML staking requirement in place.โฃ
โฃ
๐๐ฌ๐ฌ๐ฎ๐๐ฌ ๐๐ง๐๐จ๐ฎ๐ง๐ญ๐๐ซ๐๐ ๐๐ง๐ ๐ ๐ข๐ฑ๐๐โฃ
Running the internal testnet surfaced a few expected infrastructure issues which the team identified and resolved:โฃ
โฃ
- Validator nodes were getting stuck in "Waiting" and not flipping to "Active". Traced to a session rotation timing issue and fixed.โฃ
โฃ
- AlertManager was firing false alerts during normal session rotations. Monitoring thresholds have been tuned to correctly distinguish a genuine halt from expected rotation gaps.โฃ
โฃ
- Nodes were dropping peers and not reconnecting without a manual restart. Peer discovery has been stabilised and nodes now recover automatically.โฃ
โฃ
- The testnet faucet was running low on pCHML without any alert. A refill threshold and notification have been put in place so onboarding is never blocked.โฃ
โฃ
๐๐ก๐๐ญ'๐ฌ ๐๐๐ฑ๐ญโฃ
With infrastructure stable and fixes in, the team will now build and test the testnet APK. This will go through multiple rounds of iteration, similar to how the devnet APK was tested across many builds, until the experience is stable enough to release to the community.
Assign a task from your phone, turn your attention to something else, and come back to finished work on your computer.
Tell Claude once to scan your email every morning or pull a report every Friday, and it handles it from there.
The next wave isnโt AI.
It isnโt crypto either.
Itโs AI agents operating on-chain.
Autonomous.
Permissionless.
Always-on economic actors.
Most people still think this is years away.
Itโs not.
#AI#CryptoUpdates#web3โโ#AIAgents#OnchainAI#DeFi#Blockchain
Vibe coding isn't going away.
It's only getting better, more powerful, and more popular. ๐
A directory of the 29 top vibe coding tools real builders are using today (bookmark this thread):
Data wrangling, in essence, refers to the meticulous process of transforming and organizing raw data into a more accessible and interpretable format.
Read about useful data analysis expressions in this blog:
https://t.co/nfyZNnsLKu
#datawrangling#dataanalysis#datascience