clawzembic

To audit OpenClaw deployments and reduce token usage, run the lean-audit.sh script to evaluate six categories: context injection, cron health, session bloat, config health, skill bloat, and transcript size. It generates actionable remediation guidance to optimize efficiency.

OpenClaw session bloat occurs when deployments accumulate token waste, oversized transcripts, and misconfigurations. A lighthouse-style efficiency audit quantifies this waste across six critical categories and provides prioritized recommendations to shrink tokens and restore performance.

Yes, you can audit remote OpenClaw installations via SSH by executing bash skills/clawzembic/lean-audit.sh --remote user@host. This remote auditing support assesses distributed deployments without requiring you to install any agents on the target machine.

To run a self-hosted OpenClaw efficiency audit, you need Python 3.8+ standard library and the lean-audit.sh script. Execute it locally for your installation to identify inefficiencies and opportunities to reduce token usage across six categories.

The best way to check cron health and context injection is using a lighthouse-style efficiency audit. It evaluates these specific categories alongside session bloat, config health, skill bloat, and transcript size, surfacing concrete improvements to optimize your deployment.

No, the OpenClaw audit script does not require installing agents on remote machines. It supports remote auditing natively via SSH, allowing you to assess distributed deployments directly by passing the --remote user@host flag to the lean-audit.sh script.