Report security issues privately — do not open a public issue.

Use either channel:

• a private security advisory on GitHub, or
• email to security@intelligo.dev.

Include:

• what the issue is and which package or route it affects,
• the steps or request needed to reproduce it,
• what an attacker gets from it (data read, data written, access gained).

You will get an acknowledgement within three working days. If a report is valid you will be told when a fix ships and, unless you ask otherwise, credited in the release notes.

The published @intelligo-dev/* packages and the reference application. The parts most worth your attention, because getting them wrong is how a tenant sees another tenant's data or a customer is charged wrongly:

• tenant isolation — any query reachable from a request that is not scoped by workspaceId, or by userId where the data is user-private;
• authorization — a platform-admin surface reachable through a workspace role, or a workspace role escalating itself;
• the execution boundary — a way to consume model tokens without an execution record, to settle the same execution twice, or to keep a credit hold open indefinitely;
• impersonation — acting as a user without an audit event, or a session that outlives its 30-minute cap;
• webhooks — accepting a Stripe event without a valid signature, or replaying one.

Findings that require a platform administrator to already be compromised, denial of service through ordinary volume, missing hardening headers with no demonstrated impact, and reports produced solely by an automated scanner with no working reproduction.

The latest minor release of each published @intelligo-dev/* package receives security fixes. Prereleases (beta, rc) are not supported once a stable release supersedes them.