IxInfra

4 ms·

1.
by IxInfra 5mo ago
I've found that this works for smaller projects but starts breaking down when you're dealing with large-scale codebases or systems that change frequently.
2.
by IxInfra 5mo ago
Let us know if you have any feedback. We welcome contributors and are working every day to improve based on what our users want. GitHub = https://github.com/ix-infrastructure/Ix Open source & free to use
3.
Show HN: Ix. Sick of rebuilding context every session. We mapped it instead(github.com)
2 points by IxInfra 5mo ago|1 comments
4.
Ask HN: How do you stop playing 20 questions with your AI coding tools
2 points by IxInfra 5mo ago|2 comments
5.
by IxInfra 5mo ago
https://github.com/ix-infrastructure/Ix sorry heres the link.
6.
Ask HN: Deterministic codebase maps vs. LLM inferred knowledge graphs?
2 points by IxInfra 5mo ago|1 comments
7.
by IxInfra 6mo ago
Thank you for your answer.
8.
by IxInfra 6mo ago
it was a genuine question...
9.
by IxInfra 6mo ago
Interesting. Where are you getting the hooks, prompts, and skills sourced from? Are these custom-built or pulled from an existing framework?
10.
by IxInfra 6mo ago
This is really interesting, especially the chunking and parallel Haiku approach. Curious how it holds up as note volume grows. At some point you're still doing N relevance checks per tool call. do you hit a scaling limit there or does
11.
by IxInfra 6mo ago
the distribution muscle being a separate skill is so underrated. most technical founders assume if you build something good it finds its own audience. it almost never does. sounds like you're learning the hard way which is probably the
12.
by IxInfra 6mo ago
the duct tape framing is fair but the deeper issue is the model has no persistent understanding of the system it's working in. each generation starts from scratch with no memory of prior context or architectural decisions. that's
13.
Ix – We kept rebuilding context every session so we mapped the system instead(github.com)
5 points by IxInfra 6mo ago|1 comments
14.
by IxInfra 6mo ago
Hey HN, I'm Tanner, one of the founders of Ix. Every time we started a new AI session, the agent had no idea what system it was working in. Not just losing conversation, losing the entire architecture. We kept throwing more context at
15.
by IxInfra 6mo ago
The 0.92 semantic similarity threshold is interesting. curious how you landed on that number? Thinking too low and you risk false cache hits but too high and you lose most of the semantic benefit. Did you run experiments or is it tunable pe