What the platform is for
Myndstack is a cognitive-infrastructure platform: hosted models, agent orchestration, evaluation, and the data plane underneath. It is designed for teams building products for their own customers, under their own name. It is not a general-purpose hosting or storage service, and it is not a proxy for reselling raw model access.
Prohibited content and use
Do not use the platform to generate, host, or distribute: child sexual abuse material or any content that sexualises a minor; content that promotes terrorism, mass violence, or genocide; targeted harassment of a real person; unlawful surveillance or stalking; deceptive impersonation of a real person or organisation; content designed to defraud, deceive, or manipulate voters or public health.
Do not use the platform to generate or refine malware, credential-stuffing lists, phishing kits, or working exploits against systems you do not own or have written authorisation to test.
Do not use the platform to make automated decisions with legal or similarly significant effect on a person — for example credit, housing, employment, insurance, or benefits — without a documented human review step and appropriate legal basis under applicable law.
Model and data misuse
Do not attempt to extract another tenant's data, prompts, or model artefacts. Do not use our models or outputs to train a competing foundation model. Do not deliberately bypass safety filters, rate limits, or content classifiers we run on your behalf. Where you deploy your own model behind our runtime, you remain responsible for the outputs it produces.
Fair-use limits
Requests per second, storage, egress, and compute are provisioned per tier. Sustained use materially above the tier you pay for — usually because of automation, retries, or a runaway job — will trigger a conversation about moving to a higher tier or a custom quota. We will always contact you before throttling, unless we are protecting the platform from an active abuse event.
Security testing
Test your own tenant. Do not test another customer's. Do not run denial-of-service, brute-force credential, or resource-exhaustion tests against the platform without a written, dated authorisation from us. Report anything you find through the process on the security page.
Enforcement
For a first, non-critical breach we will contact you and give you a reasonable window to fix it. For a critical breach — abuse of a person, unlawful content, an active security threat, or a court order — we may suspend or terminate access immediately. Where required by law we preserve and hand over relevant records; where not required, we don't.
Reporting abuse
To report a violation of this policy on Myndstack, email hello@myndstack.io with the subject line "Abuse" and include enough detail for us to reproduce or verify the report.
Questions about this page? Get in touch.