Every product is self-assessed against a written risk-rating standard and a product-standards checklist, then reviewed before it enters the launch sequence. Ratings are drafts for human review — not certifications.
Each product passes a delegated completion review (G10) before its launch position is recorded. Prices are owner-confirmed at review.
A written model scores each product low / medium / high by whether it touches permissions, payments, licensing, incidents, or security.
Every kit meets a product-standards checklist: complete metadata, worked examples, integration pack, and rollback guide.
Canonical public pages use explicit index,follow metadata and canonical URLs. Legal, diagnostic, noncanonical, and entitlement-scoped pages remain noindex. Public agent discovery stays free and crawlable.
Disclosure via support@signalharness.ai (see /.well-known/security.txt).
Ratings are self-assessments prepared for human review, not third-party certifications.