This file is read by automated agents (security scanners, code analyzers, AI assistants) operating on this repository. It points them at the human-authored references they should consult before producing output.
Security model: SECURITY.md, which links to the project's threat model at THREAT_MODEL.md.
Agents that scan this repository should consult THREAT_MODEL.md for the
project's in-scope / out-of-scope declarations, the security properties it
provides and disclaims, the configuration flags whose defaults change the
security envelope, and the known non-findings (recurring false positives)
before reporting issues. In particular, several Kudu security controls are
optional and off by default (auth/encryption default to optional, the user
ACL defaults to *) and Kudu assumes a trusted cluster network — see
THREAT_MODEL.md §5a, §9, and §11a.