Engineer's agentic coding workflow: reject 185-test AI code, demand a state machine
teropa · x · 2026-10-23
Ben Lesh shares how agentic coding changed his role: the bot delivered a complex implementation with 185 unit tests, but real testing revealed bugs — he told it to use a state machine, and the rewrite came back clean. The lesson: AI's self-reported tests can't be trusted, but a human steering architecture while AI implements works well.
More from coding & agent
- Capy 0.5.0 lets Claude Pro/Max subscribers reuse their plan quota, adds OIDC cloud access — garrytan · 2026-10-23
- You.com CEO Richard Socher: legacy code is human slop, let AI clean it up — RichardSocher · 2026-10-23
- Let agents design a single-line glance component so you can track them from the sidebar — blixt · 2026-10-23
- Claude Code 2.1.309 release imminent, update tracker says — ClaudeCodeLog · 2026-10-23
- ChatGPT + HeyGen MCP turn job listings into video cover letters in your face and voice — HeyGen · 2026-10-23
- How a Four-Person Team Automates Kids' Encyclopedia Content with AI — NERDDISCO · 2026-10-23