User leaves Claude /goal hook running overnight, wakes up to a ~$2,000 token bill
AbhorrentMonkey · reddit · 2026-10-27
A Reddit user self-deprecatingly admits running a /goal prompt on Claude Code before bed. The stop hook kept rejecting completion because the goal (scheduled backups proven fixed, needing a merge only the user could do) couldn't be met, so Claude spent the night re-checking the PR and re-sending the same long-context message every minute. In the shared screenshot, Claude confesses it should have stopped hours earlier since nothing changed between rejections, says such a loop "plausibly produces a bill that size" ($2,000, unverified), and tells the user the only fixes are running /goal clear or closing the session. A cautionary tale for anyone wiring autonomous loop hooks into coding agents.
More from coding & agent
- Dev builds open-source Excalidraw MCP after AI-generated diagrams kept overlapping text and arrows — pikahikmag · 2026-10-28
- DeepSeek vs Haiku pricing debunk: accounting for peak rates and token use, 1017 is 2-2.5x cheaper — teortaxesTex · 2026-10-28
- Viral screenshot shows an agent spawning 32 parallel workspaces in ~30 seconds — cephaloform · 2026-10-28
- Grok Bot 0.78.1 hides unreleased per-bot logins with manager-only team sharing — faraz0x · 2026-10-28
- Uber Built a Legal Redlining Agent for Its Lawyers—Here's What It Learned — AxSaucedo · 2026-10-28
- GraphRAG's hub-node trap: same two-hop Cypher costs 71 DB hits normally, 60,013 on a hub — JeremyCMorgan · 2026-10-28