Terms of Service

Last updated: 5 May 2026

These Terms of Service ("Terms") govern your use of CovenAI's website, Agent Analytics platform, scan, monitoring, methodology rubric, developer endpoints, and MCP server. By using any of these surfaces, you agree to these Terms.

1. The Service

CovenAI provides an AI traffic intelligence platform — Agent Analytics — that measures how visible, citable, and structurally ready a website is to autonomous AI agents. The platform scores sites against our public methodology rubric (currently version 2.0.1), covering Structure, Citability, Discoverability, Agent Interface, and Transactability. The platform includes:

• Free scan — a public, no-account scan of any URL against the rubric.
• Agent Analytics dashboard — real-time and historical visibility into AI agent traffic, identity classification, citability scoring, and intelligence briefs.
• Agent Diagnostic Layer (ADL) — our proprietary intelligence layer that identifies, classifies, and scores AI agent traffic using a corpus of signatures and operator profiles. ADL operates within Agent Analytics and is not separately licensed.
• Monitoring — registered sites are scanned on a recurring schedule with score history and regression alerts.
• Methodology — the rubric, scoring weights, and grading guidance are published openly.
• Developer endpoints — public and authenticated API endpoints for programmatic scans, monitoring data, and ADL-powered analysis.
• MCP server — a Model Context Protocol server allowing autonomous AI agents to discover and use the platform.

All outputs are generated automatically by software and AI tools. CovenAI also operates Coven-Citability-Bot, our own web agent, which visits submitted URLs to perform scan and monitoring analysis.

2. Free Public Beta

CovenAI is currently in free public beta. The Agent Analytics platform, free scan, monitoring, methodology rubric, developer endpoints, and MCP server are all available at no cost. Paid tiers are not active. Beta services are provided as-is and without warranty, with no service level agreement (SLA), no uptime commitment, and no guaranteed feature stability. We may modify, suspend, or withdraw any beta feature at any time without notice. Reasonable usage limits apply to keep the service available to everyone.

3. We Do Not Transact

CovenAI is a measurement and intelligence platform. We do not currently process or facilitate transactions on behalf of users or autonomous AI agents. We have built infrastructure to support agent-initiated payments — specifically, support for the x402 payment protocol with settlement in USDC on the Base network — but this capability is not active. We will only activate it if and when the regulatory environment provides sufficient clarity for CovenAI to do so responsibly. These Terms will be updated, and a separate transaction-specific addendum may be published, before any activation. Until then, no provision of these Terms should be read as an offer to act as a payment processor, money transmitter, custodian, or financial intermediary.

4. Open-Source Components and Methodology

CovenAI releases certain components under the Apache 2.0 licence, including the scoring methodology rubric structure, the agent-policy.json specification format, the /offers JSON Schema, OpenAPI spec additions, and Agent Card format extensions for A2A compatibility. These are released freely so the broader ecosystem can benefit from shared standards. The methodology document and accompanying llms.txt are made available for use under fair, attributed terms: you may reference, quote, and build on the methodology provided you (a) attribute CovenAI clearly with a link back to the published document, and (b) do not misrepresent your own implementation as an official CovenAI score. The Agent Analytics software, ADL intelligence layer, dashboards, Identity Graph, monitoring infrastructure, developer endpoints, and MCP server that implement the methodology remain our proprietary tooling and are governed by the rest of these Terms.

5. Acceptable Use

You may only submit URLs you own or have authorisation to analyse. You must not:

• Submit URLs of third-party sites without their permission
• Use scan results or other outputs for any unlawful purpose
• Resell or redistribute paid outputs (when paid tiers are active) without written permission
• Attempt to reverse-engineer, scrape at scale, or circumvent rate limits or any billing mechanism
• Use the platform to harm, harass, or defraud other users or third parties

6. Developer Endpoints and API Use

If you use our public or authenticated developer endpoints:

• Public endpoints — anonymous endpoints (for example, the public scan endpoint) are subject to per-IP and per-origin rate limits. They are intended for individual use and product integrations, not bulk harvesting.
• API keys — once authenticated keys are issued, your key is personal to your account. You must keep it confidential and must not share it with third parties. You are responsible for all usage under your key.
• Rate limits — API requests are subject to rate limits which will be communicated to you at the time of access. Exceeding limits may result in temporary throttling or suspension.
• Fair use — you must not use any endpoint in a way that unreasonably burdens our infrastructure, circumvents billing systems (when active), or facilitates bulk data harvesting. We may suspend or terminate access where fair use is violated.

7. Agent-Initiated Use

CovenAI is built for the agentic web. We deliberately publish machine-readable surfaces that allow autonomous AI agents to discover and use the platform, including the MCP server, the agent-policy.json document, and public API endpoints. CovenAI also operates its own agent, Coven-Citability-Bot, which visits URLs submitted by users to perform scan and monitoring analysis; operators of those sites can find Coven-Citability-Bot's identity and policy at /agent-policy.json. The following terms apply to third-party agent-initiated activity:

• Equivalence — agent-initiated requests are subject to the same Acceptable Use rules and rate limits as human-initiated requests. There is no separate "agent loophole".
• Authorisation and accountability — if you operate an agent that uses an account or API key associated with you, you are responsible for that agent's activity, including any URLs it submits. You remain the data controller for any content the agent submits.
• Anonymous agent traffic — agents using public anonymous endpoints are subject to the same per-IP and per-origin limits as any anonymous client. Operators must respect any robots.txt, agent-policy.json, or rate-limit signals we publish.
• Identification and disclosure — agents should identify themselves clearly via User-Agent and follow normal good-citizen practices. We may suspend or block agents that misrepresent themselves, ignore published policies, or generate abusive traffic.
• No agent transactions — agent-initiated payments are not currently supported (see §3). Any future activation will be governed by terms published at that time.

8. Paid Tiers (When Activated)

Paid tiers for CovenAI's Agent Analytics platform (scan, monitoring, developer access, ADL data) are not active during free public beta. The following describes how paid tiers will operate when they activate. We will publish a notice on the website and update these Terms at the time of activation; the version of these Terms in force at the time of your purchase will apply to that purchase.

• Pricing — list prices will be published on the pricing page in GBP. Pricing tiers may include monthly subscriptions and one-off purchases.
• Payment method — payment method, currency, and settlement details will be set out at the point of purchase.
• Recurring subscriptions — recurring subscription cadence and minimum-term provisions, if any, will be set out at the point of purchase.
• Refunds — refund eligibility will be set out at the point of purchase. Where outputs have already been generated and delivered, refunds may be limited; we will investigate technical failures in good faith and re-deliver or refund at our discretion.
• Cancellation — for any recurring tier, you will be able to cancel through your account or by emailing hello@covenai.io. Cancellation takes effect at the end of the paid period.

9. Nature of Outputs

Scan, monitoring, ADL intelligence, and methodology outputs are generated automatically and are provided for informational purposes only. They represent measurement at a point in time against our published rubric. They should not be treated as professional legal, financial, or technical advice. Search engine rankings, agent behaviours, conversion outcomes, and business results are affected by many factors beyond our control. CovenAI accepts no liability for decisions made based on output content.

10. Intellectual Property

Outputs generated for you are licensed for your personal or business use. You may share extracts internally and reference your own scores externally with attribution. Open-source components are licensed under Apache 2.0 as described in §4. The CovenAI software, ADL intelligence layer, brand, dashboards, Identity Graph, and proprietary tooling remain our property.

11. Service Availability

We aim to keep the service available but do not guarantee uninterrupted operation, particularly during beta. Planned maintenance, third-party outages, and circumstances beyond our control may affect availability. We do not offer SLA credits for downtime during beta. If a paid SLA becomes available with paid tiers, it will be set out at the point of purchase.

12. Account Suspension

We reserve the right to suspend or terminate your account, API key, or agent access if we reasonably believe you have: violated these Terms (including the Acceptable Use policy or Agent-Initiated Use clause); engaged in fraudulent, abusive, or harmful behaviour; failed to pay any amounts due (once paid tiers are active); or caused material harm to other users or to our infrastructure. Where possible we will notify you before suspension. In cases of serious or repeated abuse, suspension may be immediate and without notice.

13. Force Majeure

We will not be liable for any failure or delay in performing our obligations where such failure or delay results from causes beyond our reasonable control, including but not limited to: acts of God, flood, fire, earthquake, pandemic, war, terrorism, industrial disputes, power failures, internet outages, or failures of third-party service providers. If such an event continues for more than 30 days, either party may terminate the affected service by written notice.

14. Limitation of Liability

To the maximum extent permitted by law, CovenAI shall not be liable for any indirect, incidental, or consequential loss arising from use of our services. For paid services, our total liability for any claim shall not exceed the amount paid by you for the relevant service in the 12 months preceding the claim. For free beta services, our total liability for any claim is capped at GBP 100.

15. Changes to These Terms

We may update these Terms from time to time. Material changes will be noted on this page with a revised date. Substantive changes affecting paid tiers, transaction features, or agent-initiated use will be notified by email where we have your address. Your continued use of the service after a change takes effect constitutes acceptance of the updated Terms.

16. Governing Law

These Terms are governed by the laws of England and Wales. Any disputes shall be subject to the exclusive jurisdiction of the courts of England and Wales.

17. Contact

Questions about these Terms: hello@covenai.io