ClawMetry shows you every AI coding agent in your organisation — what it cost, what it touched, and what it was allowed to do — across 26 runtimes, in your cloud, your data centre, or entirely air-gapped.
We are not asking you to trust a new vendor's marketing. ClawMetry has been open source since day one — you can read every line before you talk to us, and your security team can too.
Three jobs, one deployment. Most buyers arrive for the first and stay for the third.
Per-agent, per-session, per-model cost. Which repositories, which tools, which people. Runaway loops and token waste surfaced as they happen, not on next month's invoice.
Budget ceilings, tool policies, and approval queues that stop a risky action rather than logging it after the fact. Read-only by default — enforcement is something you switch on deliberately.
A tamper-evident record of what every agent did, exportable to your SIEM and mapped to the control frameworks your risk team already uses.
Collection is local in every mode. What changes is where you look at it.
We host the dashboard. Data is encrypted on your machine before it leaves and decrypted in your browser — we hold the ciphertext, not the key.
The whole thing inside your perimeter, on your infrastructure, under your own TLS and identity. No data crosses your boundary.
Offline licensing for networks with no route out. Licences verify locally against a signed key with no callback of any kind.
You are going to ask these questions in the security review anyway. Here are the headlines — the full answers, including the uncomfortable ones, are on the security page.
Enterprise engagements start at $25,000 a year.
Priced on the number of machines running agents, the deployment mode, and the support commitment you need. Smaller teams are better served by the self-serve plans on clawmetry.com — we will tell you if that is you.
Send this and you will get an immediate reply with a link to book a time. No sequence, no SDR, no drip campaign — the founder reads every one of these.