iAny runs AI on-device and also operates a hosted service (a Cloudflare Worker with R2/D1 for model mirroring, encrypted backups, Radio, and the community data collectors). We take security and user privacy seriously.
Please do not open a public GitHub issue for security vulnerabilities.
Instead, report privately by one of:
- GitHub Security Advisories — use the repository's Security tab → Report a vulnerability (preferred; lets us collaborate on a fix privately).
- Email — sengtha@gmail.com with the subject line
iAny security.
Please include:
- a description of the issue and its impact,
- steps to reproduce (proof-of-concept if possible),
- affected component (PWA, mobile app, Worker/API, a specific endpoint),
- any suggested remediation.
- We aim to acknowledge reports within 72 hours.
- We'll work with you on a fix and coordinate a disclosure timeline.
- With your consent, we're happy to credit you once a fix ships.
In scope: the iAny code in this repository and the hosted API
(/api/*, /models/*, /radio/*, backups). Especially interested in:
- authentication/authorization on admin endpoints (the
RADIO_ADMIN_TOKENguarded/api/*/admin/*routes), - issues in the encrypted backup flow,
- anything that could expose a contributor's identity in the community datasets (voice / OCR / sign), which are designed to be anonymous,
- data validation on the public upload endpoints.
Out of scope: third-party model/dataset hosts (e.g. Hugging Face), and denial-of-service via traffic volume.
All service secrets (admin tokens, storage keys) are configured as Cloudflare environment secrets and are never committed to the repository. If you believe a secret has been exposed, treat it as a vulnerability and report it privately as above.