Stories by rob-x-ai

From 300KB to 69KB per Token: How LLM Architectures Solve the KV Cache Problem

Show HN: Zsweep – Vim Motions Minesweeper with Svelte 5.0

Show HN: Unwind – I built a security proxy for AI agents on a Raspberry Pi

RegalMail – give your AI agents a real email inbox

Shown HN: Mittens for Claw – Go sandbox to safely run local AI agents

Idea Killshot: stress-test your startup idea before the market does

Elevated errors on Claude Opus 4.6

Show HN: Mengram – Open-source memory layer for AI agents

GrantAi – Deterministic memory for AI agents (exact recall, 90% fewer tokens)

Opus 4.6 1M is now the default Opus model for Claude Code users

Ask HN: Should training bottleneck detection be a product or just a feature?

MuskMeter – Minute-by-minute Musk metrics

T3 Code – TypeScript-based web and desktop GUI for "coding agents"

Show HN: Vex Runtime reliability layer that auto-corrects AI-Agent hallucination

Claude Code – Scheduled tasks (cron) added

Ask HN: 10-year dev, tried earning on the side for a year – lost $2,200

AI Has Amnesia. You're Paying. Blame the Architecture

Are there any companies who are anti-AI?

An iOS library to natively render After Effects vector animations

Nano Banana 2

1 2 3 4 5 6 7 8 9 10