Summary
Max, CTO at Hystax, wrote a Hacker News post saying his team uses Claude heavily in development and is working to make coding agents cheaper to run. They built Kompressor, an open-source tool that compresses identical tool outputs that agents repeatedly resend. In Codex tests, the team saw a real cost reduction of 22.7%. When testing with Claude Code, they found the results were completely different, and the post indicates the cache plays a central role in that difference.
AI-assisted summary based on the listed source.
What happened
Hello, HN! I’m Max, CTO at Hystax. We use Claude a lot in our development work, and we’re also trying to make coding agents cheaper to run. We built the open-source Kompressor for compressing identical tool outputs that agents constantly resend. In our Codex tests, we noticed a real cost reduction of 22.7%. But...
Why it matters
Developers and teams paying for Claude Code usage may find that their bills depend on how quickly the prompt cache expires, not only on the work performed. Tools that aim to cut agent costs may produce different savings across coding platforms. Readers should watch the full post for the specific cache behavior and any figures the authors report for Claude Code.
Signal Intelligence
Signal Strength 95%
Technical label SOURCE-BACKED
Public Interest 55
Category OPEN SOURCE
Reader Depth TECHNICAL
Event context 1 source
Signal Strength reflects source quality, relevance, freshness and evidence. Public Interest helps organize discovery; it is not proof of truth.
Public Interest components
Recognizable Entity Score 82
Practical Impact Score 28
Novelty Interest Score 70
Consequence Score 46
Curiosity Score 16
Shareability Score 64
Why this is here
VQV surfaced this signal because it is recent, relevant to AI Coding Tools, connected to Hacker News Newest.
