Terms of Service & Node Agreement

Framework & Sovereign Architecture

These Terms of Service ("Agreement") govern your access to and use of bindu.cloud (the "Platform Gateway"), the downloadable bindu.protocol client and CLI daemon, and associated APIs, orchestration interfaces, and companion sovereign native engines (local.scaler, local.build, local.create, and Indra's Net) operated by Creatives Guide Us and Sean Halls Studio ("Provider", "we", "us", or "our").

By downloading, executing, enrolling a device, or participating in the Bindu mesh, you agree to be bound by these terms. Bindu is architected on Self-Sovereignty and Anti-Hyperscaler Principles: compute workloads, storage volumes, and cryptographic keys remain in the physical custody of the respective hardware operator.

Cryptographic Identity & Key Custody

All participant authentication, node federation, and compute scheduling within Bindu are anchored by Ed25519 Sovereign Passports generated directly on participant hardware.

• Zero Private Key Exfiltration: Private keys, SSH credentials, and WireGuard preshared keys are never transmitted to, inspected by, or stored upon bindu.cloud servers.
• Local CA Certificate Model: The platform publishes a local root certificate (bindu-ca.crt) allowing participant devices to verify internal TLS endpoints within sovereign subnets (e.g. bindu.local). Trusting this certificate is voluntary and scoped exclusively to your configured domain perimeters.
• Credential Safeguarding: You are solely responsible for maintaining physical security and backups of your local keypair (~/.bindu/passport.json). Provider cannot recover lost private keys or unlock encrypted volumes if keys are misplaced.

Acceptable Use & Compute Ethics (AUP)

When participating in the cooperative silicon pool, leasing compute slots, or running inference through bindu.protocol, you agree not to:

• Execute automated denial-of-service (DoS) attacks, unauthorized port scanning, or malicious token exhaustion against peered cluster nodes.
• Attempt to break WireGuard boundary locks or access port-restricted tenant enclaves (such as Gate 6 fail-closed trading enclaves or clinical records).
• Transmit unencrypted Protected Health Information (PHI) or non-consensual biometric data across non-airgapped network segments.
• Utilize allocated contributor GPU or CPU cycles for non-consensual cryptographic mining or unauthorized surveillance tooling.

Provider reserves the right to cryptographically revoke peer admission tokens and quarantine misbehaving node identities if willful boundary violations are detected by Indra's Net telemetry monitors.

Commercial Engagements & Pay-Per-Outcome (PPO)

Commercial services contracted through the local.scaler engine operate under fixed Pay-Per-Outcome (PPO) milestone sprints:

• Clear Deliverables: Sprints are bounded by deterministic scope definitions, automated test suites, and cryptographic audit receipts rather than open-ended hourly billing.
• Payment Terms: Milestone invoices are payable upon delivery and verification. Cryptographic Merkle invoice hashes provide mutual proof of delivery and payment clearance.
• Zero Cloud Rent: Client solutions engineered through Bindu are delivered with turnkey local deployment blueprints, freeing clients from forced hyperscaler SaaS subscriptions.

Intellectual Property & Tenant Workloads

Bindu enforces strict tenant and workspace isolation:

• Client & Tenant Ownership: All proprietary code, custom models, visual storyboards, audio DSP algorithms, client contracts, and business data authored by or hosted for a tenant remain the exclusive intellectual property of that tenant or client.
• Platform Core: The Bindu control plane, local.scaler commercial algorithms, local.build AST construction pipelines, local.create diffusion orchestration engines, Indra's Net perimeter guards, and core bindu.protocol schemas are proprietary to Provider or licensed under designated open-source licenses (e.g. MIT License for specific client libraries and ViMax diffusion core).
• No Data Commingling: Tenant data processed through cluster nodes is never used to train global AI models or shared with competing tenancies.

Hardware Contributor Terms & QoS Guarantees

For operators enrolling physical workstations (such as NVIDIA GB10 workstations, Mac Studio clusters, or Spark rigs):

• Preemption & Tier Priority: Workload scheduling obeys the 4-tier QoS hierarchy (Tier 0: Interactive Operator → Tier 1: Sovereign Studio → Tier 2: Balanced Sprint → Tier 3: Background Batch). Higher priority turns preempt lower priority turns deterministically.
• Host Safety: Nodes maintain local thermal, memory, and cgroup ceilings. The protocol will never override local hardware safety limits.
• As-Is Silicon Sharing: Contributor compute is provided on a peer-to-peer cooperative basis. While QoS scheduling ensures fair weighting, Provider does not warrant uninterrupted uptime of individual distributed nodes.

Disclaimers & Limitation of Liability

TO THE MAXIMUM EXTENT PERMITTED BY APPLICABLE LAW, IN NO EVENT SHALL PROVIDER OR ITS DIRECTORS, EMPLOYEES, OR PARTNERS BE LIABLE FOR ANY INDIRECT, INCIDENTAL, SPECIAL, CONSEQUENTIAL, OR PUNITIVE DAMAGES, INCLUDING WITHOUT LIMITATION LOSS OF PROFITS, DATA, USE, GOODWILL, OR OTHER INTANGIBLE LOSSES, RESULTING FROM YOUR ACCESS TO OR USE OF THE PROTOCOL OR PLATFORM.

Governing Law & Notices

This Agreement shall be governed by and construed in accordance with the laws of the State of Utah and the United States, without regard to conflict of law provisions. Any legal action or proceeding arising under this Agreement shall be brought exclusively in state or federal courts located in Salt Lake County, Utah.

For formal legal inquiries, acceptable use notifications, or contract correspondence, contact:

Creatives Guide Us / Sean Halls Studio
Attn: Legal & Governance Team
Email: legal@creativesguide.us · sean@creativesguide.us
Web: https://bindu.cloud