We find what your GRC platform misses: shadow admins, IAM privesc, NSG holes. SOC 2 evidence. Open source. Multi-cloud. Zero data exfil.
Joined December 2009
- Tweets 1,901
- Following 46
- Followers 120
- Likes 496
84 Photos and videos
Jun 13
New podcast episode is live. We break down how NSAuditor AI Enterprise turns your AWS, Azure & GCP accounts into auditor-ready evidence — one scan, seven frameworks (SOC 2, HIPAA, NIST CSF 2.0, PCI DSS, ISO 27001, CIS v8 & GDPR Art. 32). Privacy-first, Zero Data Exfiltration.
Listen: open.spotify.com/episode/3ge…
#CyberSecurity #CloudSecurity #Compliance #Podcast
1
2
4
50
Jun 12
NSAuditor AI Enterprise 0.20.0 is live: GDPR Article 32 (Security of Processing) — the seventh framework.
One scan, seven frameworks. But read the next tweet before you read "GDPR" as "GDPR compliance." They are not the same thing. 🧵
2
2
4
29
Jun 12
Two disciplines we refused to skip:
• Every finding is substrate for the operator's "appropriate to the risk" proportionality call — never a bare PASS/FAIL.
• Fine exposure pinned to the Art. 83(4) lower tier (€10M / 2%) — never the 4% / €20M headline tier.
1
3
14
Jun 12
28 plugins · AWS / Azure / GCP · seven frameworks (SOC 2 · HIPAA · NIST CSF 2.0 · PCI DSS v4.0.1 · ISO 27001:2022 · CIS v8 · GDPR Art. 32) · Zero Data Exfiltration.
nsauditor-ai scan --host aws --compliance gdpr
→ nsauditor.com/ai/docs/gdpr/
4
29
Jun 11
4/ Security stays tight: get_findings is gated *before* any cache read — a CE/Pro caller gets the same upgrade denial as scan_cloud, never cached Enterprise findings. A committed leak-path test proves it.
1
1
8
Jun 11
5/ Paired with EE 0.19.4 (unchanged) agent-skill 0.2.10 (which teaches agents the drill-down). A green verdict you can't trust is worse than no verdict — and "the summary showed a count" looked, to an agent, exactly like "nothing there." → nsauditor.com/ai/enterprise/
1
11
Jun 11
NSAuditor AI CE 0.2.10 is live: "MCP affordance II." An audit scanner can produce a finding the reader never *counts* — and our EE 0.19.4 Desktop validation caught it live: an agent narrated "the alarm dimension came back clean" while four SQS/SNS no-alarm findings fired. 🧵
1
3
16
Jun 11
3/ And a NEW Enterprise-gated get_findings tool drills the most recent scan — filter by provider/plugin/severity/category, paginate, read the FULL untruncated text. No re-scan, no falling back to a raw cloud API (which, in that validation, hit the *production* account).
1
1
9
Jun 11
2/ The cause: the scan_cloud MCP summary itemized only CRITICAL/HIGH, so actionable MEDIUM/LOW findings were count-only — invisible to an agent that stops at the summary. Now they roll up per provider by category (count-descending, nothing hidden).
1
1
9
Jun 11
NSAuditor AI Enterprise 0.19.4 is live: "Routing-Integrity Hardening."
A scanner can fail 3 ways: what it misses, what it hides, and what it finds but never *counts* toward a compliance verdict.
0.19.x closed the first two. 0.19.4 closes the third. 🧵
2
1
3
19
Jun 11
28 plugins · AWS/Azure/GCP · six frameworks (SOC 2 · HIPAA · NIST CSF 2.0 · PCI DSS v4.0.1 · ISO 27001:2022 · CIS v8) · Zero Data Exfiltration.
A green verdict you can't trust is worse than no verdict.
→ nsauditor.com/ai/enterprise/
1
2
33
Jun 11
Prefer the LinkedIn version? Full 0.19.4 "Routing-Integrity Hardening" announcement — build-time routing guard, the honest PCI 7.2.2 partial, and SQS/SNS alarm-independence — is up on our LinkedIn page:
linkedin.com/posts/nsasoft_s…
5
Jun 11
Full write-up on LinkedIn — the routing-integrity changes in 0.19.4: linkedin.com/posts/nsasoft_s…
4