Jake Verbaten

Introducing a public Nozomio skill to test Nia on its most popular use case: providing context to coding agents. It’s a simple skill I built that doesn’t require an API key to get started, so you can try Nia before creating an account. Run 'npx skills add nozomio-labs/nia-public-skill-demo' in your terminal.

105

7 Comments

The most important investment in agentic coding velocity may have nothing to do with agents. Amplitude tripled its PR volume in six months. PR cycle time went from 5.2 hours to 44 minutes. Almost all of that came from fixing unglamorous infrastructure. 𝐂𝐈 𝐟𝐫𝐨𝐦 𝟑𝟎 𝐦𝐢𝐧𝐮𝐭𝐞𝐬 𝐭𝐨 𝟑: they swapped the JavaScript toolchain for native equivalents (tsc to tsgo, ESLint to oxlint, Prettier to oxfmt), moved to affected-only builds in Turborepo, and ran independent stages concurrently. Most tools got about 10x faster. 𝐅𝐮𝐥𝐥 𝐬𝐭𝐚𝐜𝐤 𝐮𝐩 𝐢𝐧 𝟐𝟖 𝐬𝐞𝐜𝐨𝐧𝐝𝐬: a custom CLI built on devbox starts every local service with one command, down from 3 minutes.  𝐕𝐚𝐥𝐢𝐝𝐚𝐭𝐢𝐨𝐧 𝐰𝐢𝐭𝐡 𝐧𝐨 𝐥𝐨𝐜𝐚𝐥 𝐬𝐞𝐭𝐮𝐩: add a label to a PR and get a frontend preview in under five minutes, or an ephemeral copy of the backend stack in the cloud. Once you have a full app stood up in 30 sec on every PR, testing and verification become much faster which decreases time to merge.  𝐑𝐞𝐯𝐢𝐞𝐰 𝐬𝐜𝐨𝐫𝐞𝐝 𝐛𝐲 𝐫𝐢𝐬𝐤: a model rates each code change on size, scope, test coverage, and API surface, and low-risk changes merge with no human. Cursor's Bugbot reviews every PR, and bug reports fell 55% even as PR volume tripled. Once CI dropped under five minutes, engineers stopped working locally. Remote is the new dev environment.

27

2 Comments

TIMING ASYMMETRY Displacement does not happen evenly. It happens in sequence. First: contractors. Then: new hires. Then: junior roles. Then: middle layers. Finally: senior roles: last, but hardest when it arrives. By the time senior roles are affected, escape velocity is gone. Layoffs are not the signal. They are the lagging indicator. By the time displacement is visible, retraining is already obsolete. — GRACIQ Institute #TimingAsymmetry #AIJobs #Automation #FutureOfWork #LaborReset #WorkforceShift #SystemsRisk #EconomicTransition #GRACIQ

1

Your coding agent already knows what a reservation is. What a patient is. What an invoice is. It does not know what your ProcessingContextOptions is. When your codebase speaks in domain language, the AI has a massive foundation of real-world knowledge to draw on. When it's full of implementation jargon, the AI is flying blind — and it generates more jargon next time. I got frustrated enough with this cycle that I wrote a custom TypeScript linter to catch the junk types as they appeared. And that linter kept pulling me back to Eric Evans' ubiquitous language and Martin Fowler's refactoring with fresh eyes. It turns out the case for domain-driven design just got a new argument that has nothing to do with team communication. Wrote up what I've learned so far. [link in comments]

26

5 Comments

Anthropic just shipped Claude Cowork, basically “Claude Code for non-dev work”: a desktop agent you point at a folder and it can plan and autonomously act on your files. Three quick takeaways for builders and product folks: 1. Powerful, immediate productivity non-technical users can now run multi-step automations (extract receipts → CSV, reorganize Downloads, draft reports) without writing scripts. 2. It was largely assembled by Claude Code itself Anthropic used agentic workflows to build Cowork fast, which is a notable signal about how product dev cycles may change. 3. Control & safety are the real product problem, Cowork can read/modify files in a designated folder. That’s useful and risky. Wrong instructions or weak sandboxing = accidental deletes, data leaks, or prompt-injection surprises. If you’re experimenting, run a tiny pilot: • Create a dedicated test folder (never point it at your home/production docs). • Require a pre-execution plan and manual approval for any destructive steps. • Version or snapshot everything (Git, Time Machine, simple copies). • Log prompts + plan + outputs for audit and debugging.

6