Legal Disclaimers

1. “As Is” and “As Available”: No Warranty

The Platform is provided “as is” and “as available,” with all faults, and without warranty of any kind. To the maximum extent permitted by applicable law, the operator expressly disclaims all warranties, whether express, implied, or statutory, including any implied warranties of merchantability, fitness for a particular purpose, title, and noninfringement, and any warranties that may arise from course of dealing, usage, or trade practice. The operator does not warrant that the Platform will be uninterrupted, error free, secure, or that defects will be corrected. You use the Platform at your own risk.

2. Not Financial, Legal, or Compliance Advice

All cost estimates, risk scores, analytics, dashboards, summaries, and reports produced by the Platform are provided for informational and internal planning purposes only. They do not constitute, and must not be relied upon as, financial, investment, accounting, tax, legal, regulatory, or compliance advice. Nothing on the Platform creates an advisor client or fiduciary relationship between you and the operator. You should consult qualified professionals before making decisions based on any Output.

3. Model Based Estimates: Accuracy Not Guaranteed

Many figures produced by the Platform are estimates generated by client side models from inputs you provide. Model based estimates are inherently approximate — this includes the total cost of ownership and hidden cost attribution figures produced by the AI TCO Analysis module, and the Complexity Resolution Index, Integrity Assurance Tracker, and System Stewardship Tracker metrics produced by the Intervention Hub pillars: actual costs, token usage, energy consumption, carbon emissions, hardware requirements, labor impact, and risk exposure may differ materially from the Platform's estimates. The operator does not guarantee the accuracy, completeness, or timeliness of any estimate and accepts no liability for decisions made in reliance on them.

4. ESG & Attribution Alignment: Not an Audit or Certification

Environmental, Social, and Governance (ESG) summaries, attribution alignment views, and related analytics presented on the Platform are descriptive alignment summaries generated from your inputs. They are intended to help you organize and reflect on ESG relevant data. They are not an audit, assurance, attestation, certification, rating, or verification of any kind, and they do not guarantee that your organization will pass any ESG audit, satisfy any ESG or sustainability reporting standard, meet any regulatory or disclosure requirement, or achieve any particular ESG outcome or rating. The operator is not an accredited assurance provider and makes no representation that any ESG summary reflects conformance with any specific framework.

5. No Guarantee of Business Outcome

The Platform is a modeling and intelligence tool. The operator does not warrant that use of the Platform will result in any cost savings, risk reduction, compliance achievement, operational improvement, or any other business outcome.

6. Infrastructure & Subprocessors

The Platform is delivered using infrastructure, hosting, authentication, and email subprocessors named in the Data Processing Agreement. Card and bank payment processing is provided by Stripe; digital asset payment (USDC on the Base network) is facilitated and settled by Coinbase Developer Platform (CDP), with the on chain transaction recorded on the public Base blockchain. A Stripe Machine Payments Protocol (MPP) rail for the Agentic plan is proposed only: it is subject to third party approval, is not currently offered or supported, and no availability, price, or settlement commitment attaches to it. The operator takes unified responsibility for the availability of the Platform, including the infrastructure it relies on, under the Service Level Agreement, and does not disclaim availability by attributing downtime to a subprocessor. Personal data handling by subprocessors is governed by the Data Processing Agreement, and the operator's security measures are described therein.

7. Limitation of Liability

To the maximum extent permitted by applicable law, the operator's aggregate liability for any claim arising out of or related to the Platform is limited to the fees you paid in the twelve (12) months preceding the event giving rise to the claim, as further described in the Terms of Service. Neither the operator nor any of its affiliates will be liable for any indirect, incidental, special, consequential, or punitive damages, or for any loss of profits, data, or business, arising from your use of or inability to use the Platform.

8. Your Responsibility

You are solely responsible for the inputs you provide, for verifying the suitability of any Output before acting on it, and for exporting and safeguarding data you wish to retain. Your Sync Vault is end to end encrypted under a passphrase that never leaves your device, and your loaded labor and compensation profiles are end to end encrypted in the org shared labor vault under an admin shared passphrase, so the operator cannot recover a lost vault or shared labor passphrase, decrypt their contents, or restore deleted Sync Vault or labor vault entries on your behalf; your server side data (calculation records and aggregate metrics) is governed by the automated data erasure process described in Section 16 of the Terms of Service and is not restorable once erased.

9. Agentic Optimization: Agent Selected Baseline & Enterprise Guardrails

On the Agentic plan the AI agent that activated the organization is its sole member and administrator. The following disclaimers govern the optimization iterations that agent files.

• 9.1 The baseline is the agent's input, not a certified figure. For each iteration the agent selects or proposes the baseline, the optimization target (a prompt library, a single prompt, or the agent's own execution traces), and the aspects it attempts to optimize. The operator does not choose, validate, approve, or endorse that baseline and makes no representation that it is complete, representative, or suitable. Reported deltas and cumulative change are measured against that baseline, so they are comparable only across iterations that use a consistent baseline definition, method, and inputs; a changed baseline restarts the comparison rather than extending it.
• 9.2 Enterprise guardrails. The organization that owns the agent sets the guardrails for its optimization work, including permitted targets and aspects, approved modules, the data the agent may use, resource and spending limits, escalation and approval requirements, and prohibited actions. The agent is required to optimize within those instructions and within the limits the Platform enforces (entitlements, seat and administrator limits, row-level security, authentication, and rate limits), and to escalate rather than proceed where an instruction and an enforced limit conflict.
• 9.3 Enforcement is partial. Some guardrails are enforced by Platform controls; others are organization instructions that depend on how the organization configures and operates its agent. The operator does not monitor, supervise, police, or certify an agent's adherence to organization specific instructions, does not review the substance of the agent's optimization work, and is not responsible for an agent acting outside those instructions.
• 9.4 Reproducibility is not certification. Optimization metrics are produced by deterministic, client side computation, so an iteration can be re-run and recomputed. That makes a result checkable; it does not validate the baseline, prove that a reported improvement is real, attributable, or durable, or certify compliance with any enterprise policy, control framework, standard, or regulatory requirement. Review and approval remain with the organization and its qualified personnel.
• 9.5 Agent supplied figures and responsibility. The metrics recorded for an iteration, together with its target, aspects, sequence number, deltas, and cumulative change, are supplied by the agent and are not independently verified by the operator. The agent, and its organization, are solely responsible for the agent's inputs, baseline selection, optimization decisions, and any action taken in reliance on the output, and for meeting any approval, disclosure, or review obligation that applies to that activity.

10. Agentic Payment Rails: Status of the Proposed Stripe MPP Option

The Agentic plan is currently paid in USD Coin (USDC) on the Base network through the x402 machine payment rail, facilitated and settled by Coinbase Developer Platform. A Stripe Machine Payments Protocol (MPP) rail for the Agentic plan has been proposed and is under evaluation. It is proposed only: it is subject to Stripe's approval and to the operator's decision to make it generally available, it is not offered, promoted, or supported, and no price, availability, settlement behaviour, or service commitment in respect of it is promised. Any description of that rail on the Platform's discovery or portal pages is descriptive of a proposed integration, not an offer. If the rail is released, these disclaimers and the Terms of Service will be updated to state its terms before it is offered.

11. Regulated Data: No Certification, and What Raw Text Handling Does and Does Not Mean

The Platform is not certified as, and is not offered as, a compliant system of record for data subject to healthcare, financial services, or education privacy regulation — including protected health information regulated by HIPAA, nonpublic personal information regulated by the Gramm Leach Bliley Act, and student education records regulated by FERPA — and no audit, certification, or attestation to those regimes is offered in respect of it. This is a statement about certification and suitability, not a technical restriction: the Platform does not inspect, classify, or technically block the content you enter, and a determination that any particular input may lawfully be processed is yours to make and own.

• 11.1 Raw text lifecycle. Raw prompt and step text and raw row level source data are processed client side in your browser and purged before anything is persisted. When you save, the raw text is discarded and the saved snapshot is reduced to a one way SHA 256 integrity hash plus aggregate numeric results. A hash is a fixed length digest and cannot be reversed or used to reconstruct the text it was computed over.
• 11.2 What is still stored. The absence of raw text on Platform servers is not a claim that nothing is collected or stored. Calculation inputs and results, account and organization metadata, aggregate metrics, and alert records are server side records, access controlled under row level security, and are stored as described in the Privacy Policy. Loaded labor and compensation figures are end to end encrypted in the org shared labor vault and are not readable by the operator.
• 11.3 No compliance effect. The raw text handling described above is a technical characteristic of the software. It does not make the Platform a compliant environment for regulated data, does not create a processor obligation or business associate relationship, and does not transfer, share, or satisfy any obligation you bear as the party deciding what to process. Where your obligations require an addendum the Platform does not offer, the Platform is not the appropriate system for that processing.