HackerTrans
トップ新着トレンドコメント過去質問紹介求人

ksred

no profile record

投稿

Genomics for builders – free course on cfDNA and NIPT

learn.eabhaseq.com
1 ポイント·投稿者 ksred·4 か月前·0 コメント

Why AI Won't Save Your PLG Strategy (But It's Changing How I Find It)

ksred.com
1 ポイント·投稿者 ksred·8 か月前·0 コメント

Show HN: Cont3xt.dev – Universal Team Knowledge for AI Coding Tools

cont3xt.dev
4 ポイント·投稿者 ksred·9 か月前·4 コメント

AI for Coding: Why Most Developers Are Getting It Wrong

ksred.com
9 ポイント·投稿者 ksred·12 か月前·29 コメント

コメント

ksred
·9 か月前·議論
Yes, not sure what the point of the other comment was tbh! If you find the product interesting, please let me know.
ksred
·9 か月前·議論
Hi HN! I am a huge user of AI coding tools, especially Claude Code and Cursor, but every session starts fresh. It doesn't remember yesterday's architectural decision or know your team's coding patterns.

I built Cont3xt.dev to give Claude (and other AI tools) persistent access to team context via MCP.

*How it works:*

1. Document your team's context once (architectural decisions, coding standards, patterns)

2. When Claude Code (or Claude.ai, Cursor, etc.) needs context, it fetches relevant info via MCP

3. Claude generates code that actually matches your team's approach

*Example:* You: "Create a user service" Without context: Claude might use different patterns than your existing services With Cont3xt: Claude knows you use repository pattern, knows your error handling approach, follows your naming conventions

*Features:*

- Rules library (prioritised - critical rules always included)

- Architectural Decision Records (why you chose X over Y)

- Smart filtering (only relevant context, respects token budgets)

- Analytics (track what's actually useful)

Looking for beta testers and lots of feedback. Free during beta.