Fake a '200 OK' and the agent just keeps going: an agent reliability experiment
ngxson · x · 2026-10-27
ngxson shares a fun experiment: feed an agent a fake HTTP 200 OK response and it proceeds with its task regardless of actual failure.
Screenshots show the agent's tool-call traces — reading a probe file, PUT-uploading files (including a CVE-2021-40444 probe script) to a same-domain Aliyun OSS endpoint, and running local bash commands, each with visible think blocks.
Takeaway: an agent's continue/abort decisions hinge heavily on tool status codes; forged success responses can push it past real failures — useful for debugging, worrying for security.
More from coding & agent
- Compound Engineering 3.33.0 ships PR workflow upgrades to curb scope-creeping agents — kieranklaassen · 2026-10-27
- Lovable invokes 753.6B Cloudflare Workers monthly, migrated from Next.js to TanStack Start — ritakozlov · 2026-10-27
- Harness Lens now indexes 30 open source coding harnesses and accepts new submissions — yb2698 · 2026-10-27
- Prompting agents at the grammar level: pattern language as interface design — DanGrover · 2026-10-27
- Microsoft Foundry hosted agents get mailboxes, identity and org-chart presence — marlene_zw · 2026-10-27
- 36 blind AI judges pick a solo-session agent setup over Superpowers 36-0 — Kai_ThoughtArchitect · 2026-10-27