3 个帖子 - 3 位参与者 阅读完整话题
Article URL: https://www.rebuildtheinternet.com Comments URL: https://news.ycombinator.com/item?id=47778128 Points: 1 # Comments: 0
Stop playing 20 questions with your AI every session. Every time you start a new session your agent has no idea what system it's working in. Not just losing conversation history, losing the entire architecture. We tried throwing more context at it. Didn't help. The problem wasn't retrieval. There was no map. So we built one. Ix builds a structured architectural map of your codebase deterministically from the actual code. No LLM inferring structure, no guessed relationships, nothing to drift as your code changes. ix map . runs in under 90 seconds. We've been using it ourselves and the difference in how AI agents reason about our system is pretty significant. Context that actually persists. Fewer redundant questions. Better answers. Two weeks in, 190 stars, plugins live for Claude Code, Codex, OpenCode, and Gemini CLI. Homebrew submission is in review. v0.6 coming soon! semantic retrieval, raw file support, and we hit our highest LongMemEval score yet. Open source. Good first issues labeled in the repo. Happy to answer questions or get into the technical details of how the map is built. Comments URL: https://news.ycombinator.com/item?id=47755420 Points: 2 # Comments: 1
I did a dumb thing by crawling millions of pages to find all the pricing pages I could. Then I fed all of them to ~50 LLMs to see how good or bad they all did. Then I dumped it all on a page. Just because. Here's a post on how I did this: https://pricepage.lol/how-i-built-pricepage-lol Comments URL: https://news.ycombinator.com/item?id=47740358 Points: 3 # Comments: 1