NullSec-RedTeam-AI v7.1.0 — Authorization-first foundation

Highlights

• Introduces campaign-scoped policy decisions, durable audit events, idempotency, leases, cancellation, and recovery-aware jobs.
• Makes sandbox-required execution fail closed with immutable image and least-privilege runner controls.
• Adds bounded AURA workflow contracts and transparent AI Security Lab simulation semantics.
• Adds public documentation, architecture and quality-gate visuals, CodeQL, release configuration, citation metadata, support guidance, and contributor paths.

Validation

• 40 safe local tests passed.
• Focused control-plane coverage exceeded the 70% quality threshold.
• Ruff, focused mypy, medium/high-severity Bandit checks, and PEP 517 package builds passed.

Important

This is an early-stage prerelease for explicitly authorized, isolated environments. Read RESPONSIBLE_USE.md and SECURITY.md before deployment.