NewLevin is now the product. After login you land in Levin — one autonomous AI security engineer you talk to in plain language ('scan my-repo', 'am I SOC2 ready', 'secure my app'). The old multi-page agent dashboard is gone; Levin does it all.
NewSecurity company model — Levin works like a CISO directing departments (AppSec, Red Team, Incident Response, Threat Intel, GRC), each commanding real scanners. You watch the org work an engagement, live.
NewScoped intent — a narrow command does exactly that ('scan' runs the scanner, nothing more); broad goals ('secure my app') mobilize the whole team. No more wrong-tool surprises.
NewContext-aware triage — Levin profiles what your repo IS before scanning, then judges each finding in that context, so false positives are filtered before you see them.
NewFindings as cards — severity cards open a side panel with the CVE/CWE, vulnerable code, the fix, and why it matters for your repo. Documents (privacy policy, ToS) come back as downloadable artifacts; compliance as a control matrix.
NewExpanded coverage — IaC (Checkov), containers (hadolint/Trivy), AI/LLM-security, supply-chain (typosquats, SBOM), mobile, live cloud posture, and gated offensive tools (nmap/nuclei) for assets you own.
NewProactive & autonomous — scheduled scans, per-repo memory ('what changed since last scan?'), a false-positive learning loop, and opt-in autonomous remediation bounded by your policy.
NewKnowledge base — cross-engagement correlation: 'we've seen this SSRF 47 times', systemic-issue rollups, and posture trend over time.
NewTrust layer — every step shows its reasoning, you can stop or steer Levin mid-task, and an internal immune system defends Levin's own agents from malicious inputs.
ImprovedNative settings inside Levin — billing, API keys, alerts, integrations, cloud accounts, and autonomy, no redirects.