OpenClaw Model Usagev1.2.1

Engineering managers need to understand which teams, agents, and sessions are consuming the most LLM tokens across OpenClaw deployments. This skill reads local session JSONL logs to break down token totals and costs per agent, enabling accurate internal chargebacks. It produces both a compact CLI summary and an HTML dashboard for weekly reviews.

Individual developers running multiple OpenClaw agents want visibility into their daily model usage without relying on external tools like CodexBar. They can quickly check the current model, recent usage, and token totals from their own machine. The optional HTML dashboard gives a phone-friendly view of spending trends.

Financial or healthcare organizations using local LLM agents must document which models processed sensitive data and when. This skill provides local-first, auditable summaries of per-session and per-agent usage without sending data to third-party services. Audit teams can export JSON reports for retention.

Platform teams running fleets of OpenClaw subagents across many hosts need rollups of usage by agent and session to detect runaway loops or anomalous consumption. The subagent and rows JSON outputs feed into internal monitoring pipelines. Daily summaries help capacity planning.

Independent consultants and agencies billing clients for AI-assisted deliverables can attribute token and cost usage to specific sessions or agents. This skill generates per-session breakdowns they can attach to invoices. The dashboard provides a client-friendly artifact.

A hosted layer that aggregates local model_usage.py outputs from customer environments into a multi-tenant analytics dashboard. Customers install a lightweight agent that pushes anonymized or opted-in usage summaries. Value comes from cross-team benchmarking and alerting.

The core model_usage.py script and skill remain free and open source to drive adoption. A library of premium HTML dashboard templates, theming, and export connectors (Slack, Notion, PDF) is sold as a commercial add-on. Revenue supports ongoing maintenance.

Consultants deploy and customize the model usage skill alongside broader AI cost governance policies for enterprises. Engagements include log pipeline setup, chargeback model design, and executive dashboard configuration. Ongoing advisory retains the client.