Privacy Policy

1. Overview & Data Philosophy

TheIdeaToCode (theideatocode.com) is an independent engineering tool designed to help developers and founders turn raw product ideas into practical software architecture plans and AI coding agent prompts. We believe in minimal data footprint: we do not require user accounts, passwords, or payment cards for V1 beta use.

2. Data We Process

• Submitted Ideas & Constraints: The product ideas, budget, coding experience, and architecture constraints you submit are forwarded to our configured AI model provider (Anthropic, OpenAI, Google Gemini, or DeepSeek) exclusively to synthesize your technical build plan and handoff prompt. We do not store raw prompts or idea descriptions in our backend telemetry databases.
• Temporary Conflict Continuation Tokens: When an architectural contradiction is detected between your idea and constraints, the analyzed specification is temporarily held in our serverless Redis cache under an opaque token bound to your session for up to 15 minutes. This token is marked consumed immediately upon decision submission to prevent replay, and expires automatically.
• Cost & Performance Telemetry: To prevent infrastructure runaway costs and maintain our strict daily budgets, we log aggregate operational metrics per API call: model ID, input/output token counts, thinking token counts, latency in milliseconds, pricing schedule date, and estimated API fee. No user ideas or prompts are stored in these telemetry records.
• Client & Network Identifiers: A cookie holds a random ID for your browser (no personal data). We use it, and a hashed version of your network address, to count how many free checks you have used this month and how many build plans you made today, and to stop too many requests in a short time. These counters delete themselves after about 40 days. We never store or log raw IP addresses.
• Local Storage: Your latest synthesized build plan is saved locally in your browser's localStorage for convenience so you do not lose your work upon refreshing. You can clear this at any time by clearing your browser storage or clicking “New Build Plan”.
• Voluntary Early-Access Email: If you voluntarily join our early-access waitlist, we record your email address solely to notify you when new agent integrations or capacity open. We never sell, share, or use waitlist emails for marketing blasts or unsolicited third-party promotions. You can request immediate deletion at any time by contacting us.

3. Third-Party AI Providers

Depending on backend configuration, synthesis requests are processed via commercial API agreements with Anthropic (Claude) or OpenAI. Data sent via these commercial API tiers is subject to their standard enterprise API privacy commitments and is not used to train foundational AI models.

4. External Links & Link Mode

When using Link mode, TheIdeaToCode does not crawl, scrape, or visit external URLs. You are required to provide a text description of the referenced product.

5. Contact & Data Inquiries

If you have any questions about this privacy notice, wish to request removal of an early-access email, or have security questions, please reach out directly to the operator: