Acceptance

By accessing or using the sentinel protocol, sdk, api, dashboard, or any associated tooling (collectively, "the service"), you agree to be bound by these terms. if you are using the service on behalf of an organisation, you represent that you have authority to bind that organization to these terms. if you do not agree, do not use the service.

The service

Sentinel provides a cryptographic audit layer for ai model interactions. the service seals, hashes, and anchors records of llm prompts and responses to a content-addressed ledger. sentinel does not store the content of your prompts or responses beyond what is required to generate a verification hash. the hash is the record. the content is yours.

Accounts

You must create an account to access the service. you are responsible for maintaining the confidentiality of your credentials and for all activity that occurs under your account. you must notify us immediately at security@sentinel.dev if you suspect unauthorized access. we reserve the right to suspend accounts that violate these terms or pose a risk to the integrity of the protocol.

Acceptable use

You may not use the service to audit, seal, or anchor content that is unlawful, harmful, or designed to deceive. you may not attempt to reverse-engineer, tamper with, or circumvent the sealing or verification mechanisms of the protocol. you may not use the service to generate false audit records or to misrepresent the provenance of ai-generated content. you may not resell, sublicense, or white-label the service without a separate written agreement.

Api and sdk

Access to the sentinel api and sdk is subject to rate limits published in our documentation. we reserve the right to throttle or suspend api access that exceeds fair use thresholds or threatens protocol stability. we provide the sdk under the mit license. the api itself is proprietary. your api keys are credentials — treat them as passwords.

Data and privacy

Sentinel processes the minimum data necessary to produce a verification hash. we do not read, store, or use the semantic content of your prompts or model responses. metadata — including timestamps, model identifiers, latency measurements, and region codes — is retained for up to 90 days to support audit replay and dispute resolution. for gdpr or ccpa requests, contact privacy@sentinel.dev.

Intellectual property

The sentinel name, protocol specification, dashboard, and all associated tooling are the intellectual property of sentinel labs, inc. the open-source components of the sdk are governed by their respective licenses as declared in each repository. you retain full ownership of any content you pass through the service. we claim no rights over your prompts, responses, or the outputs of your models.

Uptime and sla

We target 99.99% uptime across all edge regions. our sla for paid plans guarantees 99.9% monthly uptime, measured at the api endpoint level. in the event of a breach of sla, eligible customers receive service credits as defined in their plan documentation. credits are the sole remedy for downtime. our published p99 seal latency target is 50ms.

Fees and payment

The free tier covers up to one million seals per month. paid plans are billed monthly or annually in advance. all fees are non-refundable except as required by law or as expressly stated in your plan. we reserve the right to change pricing with 30 days written notice. continued use after a price change constitutes acceptance of the new pricing.

Termination

You may terminate your account at any time. we may terminate or suspend your access immediately if you breach these terms, if continued access poses a risk to the protocol or to other users, or if your account is delinquent. upon termination, your right to use the service ceases. audit records already anchored to the ledger are permanent and cannot be deleted — this is a design property of the protocol, not a policy choice.