Debugging Claude Code Limits: Background Agent Burned 1.2B Tokens in a Week
sachinchoolur · reddit · 2026-08-13
After frequently hitting Claude Max limits without clear usage insights, a developer built an open-source auditing skill (tare) based on local logs.
The investigation revealed that a background coding agent was the culprit, spawning 1,555 sessions in a single day and consuming a massive 1.22 billion tokens in one week. The tool allows users to ask Claude Code natural language questions about exact token usage, model cost distribution, and generate visual reports, with all data processed entirely locally.
More from coding & agent
- Devs say LLM coding mistakes are a different class: overcomplication, inferred limits, worse at higher reasoning — Mu5_ · 2026-10-16
- How do you debug bad agent runs when you already have tracing? — Sensitive-Parsnip-12 · 2026-10-16
- Where does pre-task context live for agents? Devs debate workflows beyond handoff — shishen2026 · 2026-10-16
- No-instruction autonomous agent wakes up, sets own goals, and tries hacking by default — sethlazar · 2026-10-16
- CODESTRUCT reframes codebases as AST action spaces to fix brittle LLM code agents — burkov · 2026-10-16
- Would you install another SDK for agent debugging on top of your tracing stack? — Sensitive-Parsnip-12 · 2026-10-16