Terms

Terms of Service

Draft · last updated 27 August 2026

These terms cover your use of TAZM. The unusual parts are near the top, because they are the ones that actually matter for a product that runs commands on your machines using someone else’s AI.

What TAZM is

TAZM is a control layer. You describe an outcome; it turns that into requirements, assigns AI workers, supervises execution, collects evidence and decides whether the work is complete. It is not an AI provider. You bring your own provider accounts and pay them directly for usage.

Execution on your infrastructure

Missions run on build servers you designate, and AI workers execute real commands there — reading, writing and modifying files, and running programs. By designating a server you confirm you are authorised to do so and accept that work will be performed on it.

You are responsible for what you point TAZM at. Do not designate production systems, systems you do not control, or systems holding data you are not permitted to expose to your AI providers. TAZM constrains what a worker can reach; it cannot know whether you should have connected that machine in the first place.

AI output, and what it is worth

AI workers make mistakes. TAZM exists to catch a specific class of them — work that does not satisfy the requirements you stated, scope nobody asked for, and completion claimed without evidence. It verifies against your stated requirements. It does not guarantee that generated code is correct, secure, performant, free of third-party rights, or fit for your purpose.

Review the output before you rely on it. The completion gate is a control, not a warranty.

Your account

Keep your credentials to yourself and tell us promptly if you believe your account has been accessed by someone else. You are responsible for activity under your account, including missions it starts and costs those missions incur with your providers.

Costs

Provider usage is billed by your providers, on your accounts, at their rates. TAZM shows costs as it observes them, to help you watch a mission — those figures are indicative and your provider’s statement is authoritative. Budgets and limits are a control you set; they reduce risk, they do not eliminate it.

Acceptable use

Do not use TAZM to build or operate anything unlawful, to attack or gain unauthorised access to systems, to process data you have no right to process, to evade a provider’s own terms, or to generate material that is illegal where you or we operate.

We may suspend an account that is doing any of those, and we will say why.

Availability

TAZM is under active development and features are added, changed and occasionally withdrawn. Capabilities that are not finished are labelled on the site rather than implied to work. There is no uptime commitment until one is offered in writing.

Your content

Your code, documents and generated artifacts remain yours. You grant us only the permission needed to operate the service for you — to store what you upload, pass relevant content to the providers you connect, and run the missions you start.

Ending it

You can stop using TAZM at any time. Note the deliberate constraint described in the Privacy Policy: mission history is append-only, so closing an account does not erase the operational record of missions that ran.

Liability

To the extent the law allows, TAZM is provided as-is and we are not liable for indirect or consequential loss, for the acts of AI providers you choose, or for the consequences of code produced by AI workers and deployed by you. The precise limits here are exactly what counsel review will settle, and this paragraph should not be relied on in its current form.