Agent fleets make token waste visible

The OpenClaw token bill is not weird. It is a warning.

A recent report said an OpenClaw agent fleet logged over a million dollars of OpenAI API usage in 30 days. The extreme number is less important than the pattern: once agents run in parallel, small repo and workflow inefficiencies compound fast.

The reported OpenClaw spend is a research-scale outlier, and OpenAI reportedly covered it. Most teams will not see a seven-figure bill this week. But the mechanics are the same for a five-person engineering team using Claude Code, Codex, Cursor, or internal agents.

Agents repeatedly explore repos. They read generated files. They miss validation paths. They carry stale context. They run tool loops that look busy while burning tokens. With one agent, that is annoying. With ten agents or a shared team plan, it becomes budget and throughput risk.

Source: Tom's Hardware report on OpenClaw API usage.

Public tree scan: openclaw/openclaw

A public GitHub tree scan is not a private audit, but it shows why fleet-scale agents need repo-level controls before cost becomes visible.

Repo Shape Missing agent docs, no ignore rules, huge trees, large blobs, generated files.
Run Shape Repeated grep/read loops, unclear handoff state, tool retries, stale prompts.
Proof Shape Before/after measurement, eval harness, runbook, and one concrete fix path.

What to do before scaling agents

Need this on a private repo?

The 48-hour audit finds repo and workflow cost leaks, ships a private report, adds a CI-ready threshold gate where useful, and includes one fix path so the next agent run starts cleaner.