Anthropic's Swiss cheese model explains why passing evals isn't enough for agents
hugobowne · x · 2026-09-22
Hugo Bowne-Anders applies Anthropic's Swiss cheese model to agent quality: automated evals catch regressions on tested cases, human review catches what tests miss, and production monitoring plus user feedback reveal failures in unanticipated real usage — every layer has holes. The layers only work when they catch different failure types; shared blind spots let problems pass through everything. He's teaching a free Lightning Lesson on turning failures into useful tests.
More from coding & agent
- jev-model-router Routes Claude Code Tasks Between Models via Function Hooks — dr_cintas · 2026-09-22
- A coding agent prompt: use my dotfile, keep keys safe, test locally without help — yoobinray · 2026-09-22
- Why devs skip Tinker for posttraining: cost and owning your own code — silver__tsuki · 2026-09-22
- codebase-memory-mcp indexes codebases into a persistent knowledge graph in milliseconds — tom_doerr · 2026-09-22
- Ambion: A Collaboration Kernel Running Multiple Agent SDKs in One Shared Room — andreisavu · 2026-09-22
- Formal Verification and Chaos Testing for Multi-Agent Systems with Dafny — andreisavu · 2026-09-22