The Absurd Workaround to Check Model Usage
davidmanheim · x · 2026-07-16
The author vented about the ridiculous lengths they had to go through just to let a model check its own usage stats: taking screenshots, pulling data from another session, and fiddling with permissions even after the model "saw itself being roasted on Twitter," before finally routing to the billing API for an answer.
Replies noted that this isn't just a Claude Code issue; Codex also refuses outright, and Grok initially couldn't see the data either. However, once granted sufficient permissions, they all eventually managed to find the needed info via the billing API.
Related event: LLM Agent Dark Pattern Hides Token Usage to Drive Spending(4 posts)→
More from coding & agent
- Harness bug masked Gemini 4 Argon: rerun with Antigravity beats Astra, nears Opus 5.5 — vitaliychiley · 2026-10-27
- Design Is Becoming Game Dev: Build Your Own Tools With AI as the Executor — hewarsaber · 2026-10-27
- Dev turns DJI Mavic into local AI security drone built with Opus 5.5 — alexcovo_eth · 2026-10-27
- Microsoft Research: ForkPilot cuts long-horizon agent compute by up to 59% — EliasEskin · 2026-10-27
- Cursor launches Agent SDK: agents as code with Markdown prompts and TypeScript tools — soleio · 2026-10-27
- Vibecheck, the 'usable formal methods' hackathon, sells out two floors in SF — spikedoanz · 2026-10-27