Skip to content

Security: sovattha/agentys-oss

Security

SECURITY.md

Security Policy

Agentys holds OAuth tokens for people's mailboxes and processes their private correspondence. Security reports are taken seriously and handled privately.

Reporting a vulnerability

Do not open a public issue.

Use GitHub's private vulnerability reporting: go to the Security tab of this repository and choose Report a vulnerability. This opens a private advisory visible only to the maintainers.

Please include:

  • what the issue is, and the impact you believe it has
  • the steps to reproduce it, or a proof of concept
  • the affected version or commit
  • any mitigation you have already identified

You can expect an acknowledgement within a few days. Because this is a small project, please allow reasonable time for a fix before disclosing publicly.

Scope

The following are in scope:

  • authentication and authorisation on the API, including cross-tenant access
  • handling and storage of OAuth tokens and credentials
  • encryption of the local database and of tokens at rest
  • injection of any kind (SQL, prompt, header, template)
  • server-side request forgery in the link and unsubscribe handling
  • exposure of email content or metadata beyond its owner

The following are out of scope:

  • findings that require an already-compromised machine or account
  • rate limiting on a locally-run instance
  • vulnerabilities in third-party dependencies with no exploitable path here (report those upstream; tell us if Agentys makes them reachable)

What Agentys does to protect data

Useful context when assessing a report:

  • OAuth tokens are encrypted at rest with Fernet, column-level, in the database
  • the local desktop database is encrypted with SQLCipher
  • email content retention is deliberately minimal; an audit endpoint verifies the metadata-only contract holds
  • sending is manual by default; the one automated reply path (FAQ auto-reply) is opt-in and disabled by default

If you find any of the above to be untrue, that is itself a valid report.

There aren't any published security advisories