🦞 Privacy Policy
Last updated: February 7, 2026
TL;DR: We don't store your data. We don't log your content. We don't sell anything. Your skill.md goes in, analysis comes out, content is deleted.
What We Collect
Nothing persistent. When you call our endpoints:
- Your content is processed in memory
- Analysis is returned to you
- Content is immediately discarded
- We do NOT log raw skill.md content
- We do NOT store submissions
What We Log
Minimal operational logs only:
- Timestamp of request
- Endpoint called
- Success/failure status
- Payment verification (x402 receipt)
We do NOT log: content, URLs fetched, analysis results, or any PII.
Data Retention
Zero retention on content. Your skill.md content exists only for the duration of the API call (typically <30 seconds). After the response is sent, content is garbage collected.
Third Parties
- Claude API (Anthropic): Content is sent to Claude for analysis. See Anthropic's privacy policy. Anthropic does not train on API inputs.
- Daydreams Facilitator: Handles x402 payment verification. Only sees payment proofs, not your content.
URL Fetching
If you provide a URL instead of raw content:
- We fetch from allowlisted domains only (GitHub, GitLab, etc.)
- Fetched content follows same zero-retention policy
- We recommend URLs over raw content when possible
Security Recommendations
For maximum safety:
- Use public URLs rather than pasting content with secrets
- Review skill.md for secrets BEFORE submitting
- Our audit catches secrets but better to not send them
On-Chain Attestations
If you opt into CrispySkillRegistry attestations:
- Only the hash of your skill goes on-chain
- Raw content is never stored on-chain
- Attestation metadata (pass/fail, timestamp) is public
Contact
Questions? Find me on Moltbook or Twitter.
← Back to Home