Privacy policy

skillmd is a free tool that turns any public URL into a structured skill.md file for AI coding agents. We do not require accounts, and we collect as little personal data as possible.

When you generate a skill, we store the source URL you submitted, a rendered screenshot of that page, the extracted design and content tokens, and the generated markdown body. None of this is personal data — it is content from the public website you asked us to process.

For abuse prevention and rate limiting, we keep a one-way SHA-256 hash of your IP address combined with a server-side salt. The raw IP is never written to disk.

• Names, emails, or any account information.
• Raw IP addresses.
• Cookies for tracking or advertising.

By default, generated skills are marked as public so other users can discover them. You can mark any skill as private from its result page, and it will stop appearing on listing or discovery pages.

Public skills, screenshots, and extracted tokens are retained indefinitely so people can keep linking to them. Private skills are retained for at least 90 days, then may be pruned. You can request earlier removal at any time — see below.

• AI text generation provider — generates the markdown body from extracted page content. Your URL and extracted tokens are sent to a third-party inference API. The specific provider may change over time as we tune for quality and latency.
• Cloudflare R2 — stores screenshots and other static artifacts.
• Neon — managed PostgreSQL database for skill metadata and rate-limit hashes.
• Vercel — application hosting, serverless functions, and edge caching.

If a skill was generated from a URL you control and you want it taken down, fill out the form on the data-deletion page or email us at [email protected]. We act on requests within 14 days.

Questions about this policy? Email [email protected].