HN Explorer

shad42

↗ HN profile [ 73.0 ms ] full profile
Karma
0
Created
()
Submissions
0
  1. Automating the engineering work that isn't your product (mendral.com)
    2 points 0
  2. How we know if our agent is right (mendral.com)
    3 points 0
  3. Not update your dependencies in 2026 (mendral.com)
    1 points 0
  4. Supply chain attacks don't wait for CVEs (mendral.com)
    3 points 0
  5. We built an agent for CI despite using Claude Code for everything else (mendral.com)
    2 points 0
  6. We built an agent specialized for CI (using the 3 Claude models) (mendral.com)
    2 points 0
  7. How we know if our agent is right (mendral.com)
    2 points 0
  8. How we know if our agent is right (mendral.com)
    5 points 0
  9. The agent harness belongs outside the sandbox (mendral.com)
    182 points 52
  10. We decreased our LLM costs with Opus (mendral.com)
    106 points 14
  11. Multi-player agents don't fit in the sandbox (mendral.com)
    2 points 0
  12. We built our AI agent, for analyzing CI logs (mendral.com)
    1 points 0
  13. Same LLM, different agent: a CI debugger built on Claude (mendral.com)
    2 points 0
  14. Agent Harness: Inside vs. Outside the Sandbox (mendral.com)
    3 points 0
  15. Same LLM but different output: we built a CI specialist (mendral.com)
    1 points 0
  16. We upgraded our agent to Opus and our costs went down (mendral.com)
    2 points 0
  17. Same LLM, Different Agent: What Changes When You Specialize for CI (mendral.com)
    3 points 0
  18. We decreased our LLM costs by switching to Opus (mendral.com)
    2 points 0
  19. What CI looks like at a 100-person team (PostHog) (mendral.com)
    56 points 10
  20. We upgraded to a frontier model and our costs went down (mendral.com)
    1 points 0
  21. We gave terabytes of CI logs to an LLM (mendral.com)
    233 points 37
  22. Anatomy of a Production AI Agent (mendral.com)
    2 points 0
  23. What CI Looks Like at a 100-Person Team (mendral.com)
    3 points 0
  24. What CI looks like at a 100-person team (mendral.com)
    4 points 0
  25. Evals as Code: CI for LLMs with Dagger (dagger.io)
    2 points 1