Terms of Service

1. Agreement

By creating an account or using AgentNoah (“the Service”), you agree to these Terms of Service (“Terms”). If you do not agree, do not use the Service.

AgentNoah (“we”, “us”, “our”) is a product by AgentNoah, based in the Philippines.

2. Service Description

AgentNoah is an AI-powered code audit platform that:

• Connects to your GitHub repositories (with your explicit authorization)
• Analyzes your code using multiple AI models to identify security vulnerabilities, performance issues, and quality problems
• Generates findings with severity ratings and confidence scores
• Optionally generates automated fix suggestions via pull requests

3. Account & Access

• You must authenticate via GitHub OAuth to use the Service.
• You are responsible for maintaining the security of your GitHub account.
• You must be at least 18 years old or the age of majority in your jurisdiction.
• One person or legal entity per account. Shared accounts are not permitted.

4. Plans & Billing

Effective 2026-05-02 (v14 pricing): single Pro tier at $39/mo ($29/mo locked-forever for limited-time founding customers) plus an optional usage-based REPLACE add-on at $3 per autonomous fix delivered. Legacy three-tier ($39 Starter / $89 Pro / $229 Team) and v13 pricing both retired — customers on legacy plans retain their grandfathered subscription terms until next renewal.

REPLACE — Pay-Per-Use Add-On

REPLACE (autonomous fix mode) is opt-in per repository and billed separately on a usage basis. Pro subscription required.

• $3.00 USD per autonomous fix delivered. Billing is monthly: fixes delivered during a calendar month are tallied and charged via Paddle as one consolidated one-time charge in the first 5 days of the following month. Customers receive a single confirmation email per billing cycle.
• 1 free fix per repository (one-time lifetime trial; trial state is per-repository, not per-account; v14 default effective 2026-05-12, reduced from the v13 value of 3)
• Hard cap: 300 fixes per calendar month per account. Cap counter resets on the 1st of each month. AgentNoah will not deliver autonomous fixes once the cap is reached, regardless of repository toggle state.
• Email notifications: warning at 250 fixes (80% of cap) and cap-reached at 300 fixes.
• You may toggle REPLACE on or off per-repository at any time from Settings. Toggling off does not refund already-delivered fixes from the current period.
• REPLACE charges are billed via Paddle as one-time charges at the end of each calendar month for fixes delivered in that month.

Billing Terms

• Pro subscription is billed monthly via Paddle (our merchant of record / payment processor).
• Prices are in USD. Tax may apply based on your jurisdiction; Paddle handles tax collection and remittance.
• You may cancel at any time via the billing portal. Access continues until the end of the current billing period.
• We reserve the right to change pricing with 30 days notice to existing subscribers. Founding-customer grandfathering (limited-time founding cohort) preserves pricing indefinitely as documented at signup.

5. Refund Policy

All payments are processed by Paddle as our Merchant of Record. Paddle handles all billing, tax, and refund processing.

• 30-day money-back guarantee: If you are unsatisfied for any reason, you are entitled to a full refund within 30 days of any payment. No conditions, no exceptions, no questions asked.
• To request a refund, contact agentnoah.dev@gmail.com or use Paddle's billing portal. Refunds are typically processed within 5-10 business days.
• This refund policy is governed by Paddle's Buyer Terms. Paddle, as the Merchant of Record, is responsible for processing all refunds.

6. Your Code & Data

This is critical — please read carefully:

• We do NOT store your source code.Most operations run in YOUR IDE LLM via the BYOL MCP, so AgentNoah's servers never see your full source code during those operations. When the daemon does need to fetch source from GitHub at audit time (to render the IDE-LLM prompt or apply a REPLACE auto-fix), the file contents are processed in memory and discarded after the operation completes.
• We DO store: Audit findings (vulnerability descriptions, file paths, line numbers, severity), workspace metadata (tech stack, business context), and usage metrics.
• We NEVER use your code to train AI models. Most operations (FIND audit, FIX-via-IDE, BUILD, Discovery, Red Team Brief) run inside YOUR IDE LLMvia the BYOL MCP — Claude Code, Cursor, VS Code Copilot, or any MCP-capable IDE. AgentNoah's servers never see your full source code during these operations. Two narrow server-side exceptions: (a) Google Gemini Embedding API — embedding vectors of finding descriptions (NOT source code) for similarity search and cross-audit memory; (b) Anthropic Claude Sonnet 4.6 — used only when REPLACE autonomous-fix mode is enabled per repo, and only for the specific commit that triggered the fix. Both providers state in their API terms they do NOT use API inputs to train models.
• AI provider terms apply. By using the Service, you acknowledge that your code is processed by third-party AI providers subject to their respective terms of service and data handling policies.
• See our Privacy Policy for full details on data handling.

7. Acceptable Use

You may NOT use AgentNoah to:

• Audit repositories you do not own or have authorization to audit
• Attempt to extract, reverse-engineer, or abuse the AI audit pipeline
• Circumvent usage limits, rate limits, or authentication mechanisms
• Transmit malicious code designed to exploit the audit pipeline
• Resell or redistribute audit results as a competing service

8. Intellectual Property

• Your code is yours. We claim no ownership over your source code or repositories.
• Audit findings are yours. You own the findings generated from your audits.
• The Service is ours. The AgentNoah platform, audit pipeline, prompts, algorithms, and interface are our intellectual property.

9. Service Availability

• AgentNoah is provided “as is” without warranty of any kind.
• We do not guarantee 100% uptime. The Service runs on Google Cloud Platform (Cloud Run + Cloud SQL in asia-southeast1) behind Cloudflare. The BYOL MCP audit/build/discovery/red-team operations also depend on the availability of your IDE LLM provider (Anthropic for Claude Code, OpenAI/Anthropic for Cursor, GitHub Copilot for VS Code, etc.). Server-side AI dependencies are limited to Google (Gemini Embedding API for similarity search) and, when REPLACE autonomous-fix mode is enabled per repo, Anthropic (Claude Sonnet 4.6 for fix generation).
• We may perform maintenance, updates, or experience outages. We will attempt to notify users of planned maintenance.
• AI-generated findings are not guaranteed to be accurate, complete, or free of false positives. Always review findings before acting on them.

10. Limitation of Liability

To the maximum extent permitted by law:

• AgentNoah is not liable for any damages resulting from the use or inability to use the Service.
• AgentNoah is not liable for any security vulnerabilities that the Service fails to detect.
• AgentNoah is not liable for any damages resulting from acting on AI-generated findings.
• Our total liability shall not exceed the amount you paid us in the 12 months preceding the claim.

11. Account Termination

• By you: You may delete your account at any time from Settings. This permanently deletes all your data (findings, repos, subscriptions).
• By us: We may suspend or terminate accounts that violate these Terms, with or without notice.

12. Changes to Terms

We may update these Terms at any time. Material changes will be communicated via the email associated with your GitHub account. Continued use after changes constitutes acceptance.

13. Governing Law

These Terms are governed by the laws of the Republic of the Philippines. Disputes shall be resolved in the courts of Metro Manila, Philippines.

14. Contact

For questions, refund requests, or concerns:

• Email: agentnoah.dev@gmail.com
• GitHub: github.com/guevae2