Y Hacker News new | ask | show | jobs

user: shad42 created: 2010-09-28 karma: 652 Building https://www.mendral.com/ - prev. Dagger co-Founder; Docker 1st eng submissions: Automating the engineering work that isn't your product 2 points | 0 comments How we know if our agent is right 3 points | 0 comments 0 points | 0 comments Not update your dependencies in 2026 1 points | 0 comments Supply chain attacks don't wait for CVEs 3 points | 0 comments We built an agent for CI despite using Claude Code for everything else 2 points | 0 comments We built an agent specialized for CI (using the 3 Claude models) 2 points | 0 comments How we know if our agent is right 2 points | 0 comments How we know if our agent is right 5 points | 0 comments 0 points | 0 comments 0 points | 0 comments 0 points | 0 comments 0 points | 0 comments The agent harness belongs outside the sandbox 182 points | 121 comments 0 points | 0 comments 0 points | 0 comments 0 points | 0 comments 0 points | 0 comments 0 points | 0 comments 0 points | 0 comments 0 points | 0 comments We decreased our LLM costs with Opus 106 points | 31 comments Multi-player agents don't fit in the sandbox 2 points | 0 comments We built our AI agent, for analyzing CI logs 1 points | 0 comments Same LLM, different agent: a CI debugger built on Claude 2 points | 0 comments Agent Harness: Inside vs. Outside the Sandbox 3 points | 0 comments Same LLM but different output: we built a CI specialist 1 points | 0 comments We upgraded our agent to Opus and our costs went down 2 points | 0 comments