Token Audit

Workspace token consumption analyzer for OpenClaw agents. Scans workspace files, estimates token counts across 6 major models (GPT-4, Claude, Gemini, Llama, Mistral, Qwen), and surfaces hidden cost drivers. Identifies bloated context, redundant files, and optimization opportunities. Free, zero dependencies, no API keys.

Analyze your OpenClaw workspace token consumption and costs across multiple models.

Usage

Scan your workspace

python3 scripts/token-audit.py

Scan a specific path

python3 scripts/token-audit.py --workspace /path/to/workspace

JSON output

python3 scripts/token-audit.py --json

What It Finds

• Total token count per model (GPT-4, Claude 3, Gemini, Llama 3, Mistral, Qwen)
• Per-file token breakdown sorted by size
• Estimated costs per model at current pricing
• Files that are disproportionately large vs their utility
• Optimization recommendations

Part of Caelguard

Free, open-source agent security toolkit. See also: shellguard-scanner