alt Hacker News> and finding the best solution for every problem
Genuinely: how? Share some insights. So far I have seen anything exceeding about 2k lines of text in the context diverge. Meaning, throwing more LLM at it rapidly baloons the size at the expense of internal coherence. Things become stale, hallucinated, duplicated and outright faked. Only thing that works is constant manual intervention and pruning of the generated slop.
Not only these things are unable to find "best solution", they are unable to find any solution. Instead (especially for Opus) they seem optimized to convince user that the task is accomplished.
I crafted my own /investigate skill. Think of it as a version of /grill-me where you propose an investigation and dig into it until you come up with a spec backed by evidence.
My investigation skill can create one, two, or dozens of probes with benchmarks. Once you're confident in the solution, you can plan and author the investigation, then execute it phase by phase.
- Post https://www.minid.net/2026/7/14/how-to-automatise-with-ai
- GitHub repo with examples of skills and rules. https://github.com/meerita/monorepo-nextjs-golang-rust-pytho...