Terms of Service
These Terms of Service ("Terms") govern your use of the Canopy platform and the HookGuard Core Engine provided by HookGuard, Inc.
- Effective Date: October 26, 2026
- Last Updated: October 26, 2026
- Legal Inquiries: privacy_officer@hookguard.ai
- Security & Operations: security@hookguard.ai
1. Agreement to These Terms
These Terms constitute a legally binding agreement between HookGuard, Inc. ("Company," "Canopy," "HookGuard," "we," "us," or "our") and you or the entity you represent ("Subscriber," "MSP Partner," "User," "you").
By accessing the Canopy Control Cockpit, utilizing our APIs, connecting webhooks, or authorizing our Domain-Wide Delegation (DWD) client ID in a Google Workspace Admin Console, you agree to these Terms. If you do not agree, discontinue use immediately and revoke our DWD authorization.
We may update these Terms periodically. Continued use of the platform after changes are published constitutes acceptance of the revised Terms.
2. Definitions
- "Services" — The Canopy platform, Canopy Control Cockpit, HookGuard Core Engine, Claymore Engine, DriveHygiene, Aurelius Core, APIs, and associated webhooks hosted at hookguard.ai and canopy.hookguard.ai.
- "MSP Partner" / "Subscriber" — Managed Service Providers, consultants, or organizations managing internal or client Google Workspace domains through the platform.
- "Managed Tenant" — A distinct Google Workspace domain connected to or audited by the Services.
- "Metadata" — Technical, administrative, and permission attributes (e.g., OAuth client IDs, scope URIs, application names, file names, sizes, timestamps, and sharing flags), strictly excluding the actual substantive contents of files, emails, or messages.
- "Claymore Engine" — The subsystem executing programmatic OAuth token quarantines and revocations.
- "Aurelius Core" — The translation subsystem using AI (Gemini 2.5 Flash) to generate plain-English scope explanations and structured webhook payloads.
3. Description of Services & Advisory Disclaimer
Canopy provides automated Workspace Governance & Hygiene (WGH) across Google Workspace environments, focusing on Non-Human Identity (NHI) scope governance, OAuth monitoring, and file sharing posture.
Advisory Disclaimer: Canopy is an automated administrative tool. The Services do not constitute a formal audit, cybersecurity guarantee, penetration test, or certification of regulatory compliance (such as SOC 2, HIPAA, or ISO 27001). Risk labels and natural-language posture translations are advisory only, intended to assist qualified IT professionals, and should not be relied upon as the sole basis for security, operational, or business decisions.
4. MSP Authority & Multi-Tenant Management
When deploying Canopy across third-party client estates, you represent and warrant that:
- You possess valid legal and contractual authority from each Managed Tenant to connect their Google Workspace domain, inspect directory metadata, and execute administrative remediations.
- You are solely responsible for managing your technicians' administrative accounts and access controls within the Canopy Control Cockpit.
- HookGuard contracts exclusively with the Subscriber; we hold no direct contractual relationship with your underlying client tenants.
5. Domain-Wide Delegation & Keyless Runtime
Canopy connects to Google Workspace via Google Cloud Platform service accounts using Domain-Wide Delegation (DWD). We operate keyless infrastructure utilizing Application Default Credentials (ADC).
You maintain continuous administrative control over this connection and may revoke Canopy’s access at any time directly within the Google Workspace Admin Console (admin.google.com), which immediately stops all platform sweeps.
6. Remediation Actions & Downstream Operational Risk
Canopy provides tools (including the Claymore Engine) to suspend or revoke OAuth tokens (admin.tokens.delete) and adjust external file-sharing links.
Assumption of Risk: Programmatically severing third-party integrations, OAuth tokens, or automated workflows may disrupt downstream applications or business operations. All severance actions are initiated or authorized by the Subscriber. HookGuard, Inc. is not liable for workflow breakage, operational downtime, or third-party service disruptions resulting from token revocations or permission changes executed through the platform.
7. Generative AI Use (Aurelius Core)
Canopy uses Gemini 2.5 Flash via enterprise Google Cloud Vertex AI strictly as an ephemeral translation engine within Aurelius Core:
- Translation Only: The model translates raw, technical OAuth scope URIs into readable summaries and structured JSON for alerts and webhooks.
- No Autonomous Actions: The AI has no execution authority. All token revocations or quarantine actions are executed deterministically by a human operator or static policy via the Google Admin Directory SDK.
- No Training on Customer Data: Your data is never used to train or fine-tune public foundation models.
- Deterministic Fallback: If Vertex AI is unreachable, the system automatically falls back to static regular expressions to maintain operational visibility.
(A detailed technical and architectural whitepaper on our Aurelius Core isolation model is available upon request at privacy_officer@hookguard.ai.)
8. Intellectual Property
All software, algorithms, heuristics, interfaces, documentation, and trademarks associated with Canopy and HookGuard are owned by or licensed to HookGuard, Inc. We grant you a limited, revocable, non-exclusive, non-transferable license to use the platform for your internal business and client management purposes. You agree not to reverse engineer, decompile, scrape, or build a competing product based on our platform.
9. Subscriptions, Invoicing & Fair Use
- Billing Tiers: Services are billed based on active managed seats or wholesale tenant tiers. Paid subscriptions are billed in advance on a recurring monthly or annual basis.
- Suspension for Non-Payment: Accounts past due by more than thirty (30) days may have automated scanning and webhook dispatches suspended until resolved.
- Fair Use: We reserve the right to throttle or rate-limit accounts exhibiting abusive or non-standard API traffic that degrades multi-tenant infrastructure.
10. Third-Party Infrastructure Dependencies
Canopy relies on third-party cloud infrastructure, including Google Cloud Platform, Google Workspace APIs, and Upstash Redis. We are not liable for service interruptions, API rate limits, or downtime caused directly by upstream provider outages.
11. MSP and Provider Liability Shield
When you use the Services on behalf of your clients, HookGuard, Inc. is not a party to that commercial relationship. We do not supervise or endorse the advice, reports, or remediations you deliver. All advisory outcomes, compliance memos, and client representations are your sole responsibility, and HookGuard bears no liability for provider-client disputes.
12. Disclaimer of Warranties
The Services are provided "AS IS" and "AS AVAILABLE" without warranties of any kind, express or implied. We disclaim all warranties, including merchantability, fitness for a particular purpose, and non-infringement. We do not warrant that the Services will detect every rogue application, scope change, or file exposure.
13. Limitation of Liability
To the maximum extent permitted by law, HookGuard, Inc. and its officers, directors, and employees shall not be liable for any indirect, incidental, special, consequential, or punitive damages (including loss of profits, data, or business interruption) arising from your use of the Services or token revocations.
Liability Cap: In no event shall our total aggregate liability exceed the greater of: (a) the fees you paid to HookGuard, Inc. for the Services in the three (3) months preceding the claim, or (b) one hundred United States dollars ($100.00).
14. Indemnification
You agree to defend, indemnify, and hold harmless HookGuard, Inc. against any third-party claims, damages, or costs (including reasonable legal fees) arising from: (a) your use of the Services; (b) your violation of these Terms; or (c) any claim by a Managed Tenant regarding unauthorized domain binding or service disruption resulting from actions you took through the platform.
15. Governing Law & Dispute Resolution
These Terms are governed by the laws of the State of Delaware, without regard to conflict-of-law rules.
The parties agree to first attempt informal resolution by contacting privacy_officer@hookguard.ai. Unresolved disputes shall be resolved exclusively through confidential, binding arbitration administered by the American Arbitration Association (AAA) under its Commercial Arbitration Rules.
16. General Provisions
- Severability: If any provision is found invalid, the remaining provisions remain in full force.
- Survival: Sections regarding intellectual property, disclaimers, liability limitations, indemnification, and dispute resolution survive termination.
- Entire Agreement: These Terms, together with our Privacy Policy, constitute the entire agreement between you and HookGuard, Inc.
Privacy & Anti-Extraction Policy
This policy explains how HookGuard, Inc. ("Company," "we," "us," or "our") handles data when you use the Canopy platform and the HookGuard Core Engine (the "Services").
- Effective Date: October 26, 2026
- Last Updated: October 26, 2026
- Privacy Officer & Compliance: privacy_officer@hookguard.ai
- Security & Incident Operations: security@hookguard.ai
1. The Anti-Extraction Commitment
We do not participate in the data extraction or telemetry brokerage economy. Canopy is engineered to audit and protect Google Workspace perimeters from accidental data exposure and shadow integrations.
We do not sell personal data, we do not inspect file contents, and we do not train generative AI models on your data.
2. Data We Collect: Metadata Only
Canopy operates on a strict pure metadata model. We request the minimum Google Workspace scopes needed to audit your perimeter:
-
What We Collect:
- OAuth & Identity Metadata: Application names, client IDs, authorized scopes (URIs), authorization dates, and deterministic SHA-256 hashes of scope sets to detect scope drift.
- File System Metadata (DriveHygiene): File names, IDs, MIME types, file sizes, timestamps, MD5 hashes, and sharing permissions (such as public link flags like
anyoneWithLink). - Administrative Account Data: Names, email addresses, and login credentials of registered MSP technicians.
- Integration Settings: Configured webhook URLs and alert thresholds for downstream systems (e.g., Zapier, Make, Rewst).
-
What We NEVER Access or Ingest:
- ❌ The body or content of Google Docs, Sheets, Slides, PDFs, or Drive files.
- ❌ The content or subjects of Gmail messages, chats, or attachments.
- ❌ User passwords or corporate credential stores.
3. Google API Limited Use Compliance
Canopy's use and transfer to any other app of information received from Google APIs adheres strictly to the Google API Services User Data Policy, including the Limited Use requirements:
- We use Google Workspace API data solely to provide administrative workspace governance, hygiene metrics, and scope management.
- We never transfer this data to advertising platforms, data brokers, or information resellers.
- Humans do not read your metadata unless you give explicit, authenticated consent for technical support or where required by law.
- Data obtained via Google APIs is never used to train generalized AI or ML models.
4. Generative AI Data Protections (Aurelius Core)
Canopy uses Gemini 2.5 Flash through enterprise Google Cloud Vertex AI solely to translate technical scope strings into human-readable liability explanations:
- Ephemeral Processing: Data sent to Vertex AI consists only of cryptic scope URIs and application names. No file contents, email bodies, or personal documents are ever submitted.
- Zero Model Training: Prompts and outputs are never retained by Google to train or improve foundation models.
- Transient Memory: Model outputs are cached locally in your browser session and discarded upon closing the tab.
- No Autonomous Actions: AI outputs are purely informational. Any subsequent token revocation is executed deterministically by a human operator or static rule.
(If you require full enterprise compliance documentation regarding our AI data boundary, contact privacy_officer@hookguard.ai.)
5. Third-Party Service Providers (Sub-Processors)
We partner with cloud providers that maintain recognized security certifications (including SOC 2 Type II and ISO 27001):
| Provider | Purpose | Data Processed |
|---|---|---|
| Google Cloud Platform (GCP) | Compute (Cloud Run, Cloud Tasks), database (Firestore), and audit ledger (BigQuery) |
Metadata, scope hashes, operational logs |
| Google Cloud Vertex AI | Semantic scope translation (Gemini 2.5 Flash) | Ephemeral scope URIs & app names (Zero training) |
| Upstash Redis | Distributed worker locking & transient scan checkpoints | Ephemeral scan cursors (short TTL) |
| Firebase Auth | Technician identity and session authentication | Technician email and auth tokens |
We do not sell, rent, or trade your data to third parties.
6. Data Storage, Retention & Deletion
- Operational Data (Firestore): Active tenant mappings and current scan summaries are retained only while a tenant is actively connected.
- Audit Trail (BigQuery): Token revocations, scope changes, and security events are recorded in an append-only, cold audit ledger retained for 365 days to support compliance requirements (such as SOC 2 CC6.1), after which they are permanently purged.
- Tenant Offboarding & Right to be Forgotten: When you disconnect a domain or delete a tenant in Canopy:
- All API sweeps and task queues stop immediately.
- Operational records in Firestore and transient cache keys in Redis are permanently purged within thirty (30) days.
- Historical audit records in BigQuery expire automatically at the end of their 365-day lifecycle.
7. Data Security Safeguards
We implement technical and operational controls aligned with Google Cloud CASA Tier 2 and SOC 2 Type II:
- All data in transit is encrypted using modern TLS (1.3).
- All stored data is encrypted at rest using AES-256.
- We operate keyless serverless infrastructure using Google Application Default Credentials (ADC), eliminating static private keys.
- Multi-tenant data isolation is enforced at the database and application layers.
8. Security Incidents
In the event of a verified security incident affecting customer metadata, we will notify impacted MSP Partners via email within seventy-two (72) hours of confirmation, outlining the incident, affected metadata categories, and our remediation actions.
9. Privacy Rights & Inquiries
Depending on your jurisdiction (including under GDPR and US state privacy laws such as CCPA/CPRA), you may have the right to access, correct, or request deletion of personal administrative information we hold.
To exercise these rights, or to request our standard Data Processing Addendum (DPA) or comprehensive security whitepaper, contact our compliance office:
HookGuard, Inc.Email: privacy_officer@hookguard.ai
Portal: canopy.hookguard.ai