Show HN: Coherence – drift detector for AI-driven repos
https://github.com/fireharp/coherence
I’d be interested in feedback on the model, false positives, and whether the examples match drift people are seeing in larger repos.
Remote: Yes, remote-first or EU/global remote
Willing to relocate: No
Technologies: Python, TypeScript/React/Node.js, Django/FastAPI, Go, SQL, PostgreSQL/MySQL, BigQuery, Redis, Linux, GCP/AWS, CI/CD, Playwright, LLM workflows, MCP, E2B, evals/harnesses
Résumé/CV: https://fireharp.com/cv?access_key=hn-wants-to-be-hired-2026...
Email: [email protected]
Hi, I'm Alex. I love to build software and solve complex tasks. Looking for a team I'd spend next decade with )
https://github.com/fireharp/coherence
I’d be interested in feedback on the model, false positives, and whether the examples match drift people are seeing in larger repos.
https://docs.mumbli.app/benchmarks
It'll be interesting to see it on my coding evals as well. Can't do it yet but will try later.
I’m not a designer either. I basically took Claude Design, and practiced a little bit.
For me, as an engineer who doesn’t really do design at all, it feels pretty reasonable. Recently I would usually just take Tailwind and some decent default component set, like shadcn, and build from there. Now I can do a reasonable version myself with Claude, which is cool.
title doesn't let nuance, ofc it's not the app that's faster but the way you can use it with Groq inference for example.