nsauditor-ai-agent-skill 0.2.2 → 0.2.3

package/CHANGELOG.md

@@ -4,6 +4,12 @@ Release notes for **`nsauditor-ai-agent-skill`** — installable knowledge packa
 
 ---
 
+## 0.2.3 (2026-06-05) — Paired-release pin for EE 0.18.2 + CE 0.2.3 — scan_cloud evidence-gap visibility (end-to-end)
+
+Paired no-op bump (no agent-skill content change beyond the SKILL.md version banner; SKILL.md body + `references/plugins.md` UNCHANGED). EE 0.18.2 + CE 0.2.3 make the no-false-clean evidence-gaps the cloud plugins emit **visible through the MCP `scan_cloud` transport, end-to-end**: the CE collector renders a dedicated "Evidence gaps (unverified)" section, and a new EE CI producer-contract guarantees every cloud plugin (AWS / Azure / GCP) marks its scan-coverage gaps so they reach it (retrofitted AWS S3 1020 + Azure 1220/1221/1222 + AWS IAM 1030). EE 0.18.2 also hardens the hand-rolled source scanners against a regex-literal desync (including the ZDE read-only security meta-test, where it could have masked a mutating cloud call) and adds the proprietary `LICENSE` / EULA (now shipped in the package) + per-file copyright headers. Plugin count UNCHANGED at 28; all six coverage matrices UNCHANGED.
+
+---
+
 ## 0.2.2 (2026-06-05) — Paired-release pin for EE 0.18.1 + CE 0.2.2 — GCP false-negative hardening II + read-only enforcement
 
 Paired no-op bump (no agent-skill content change beyond the SKILL.md version banner; SKILL.md body + `references/plugins.md` UNCHANGED). EE 0.18.1 closes three more GCP false-negative defects at substrate depth on already-covered controls — **(1) plugin 1021** split-range firewall rules that cover the whole IPv4 internet without literally writing `0.0.0.0/0` now flag as the `0.0.0.0/0` CRITICAL (conservative full-coverage helper); **(2) plugin 1025** the SA-impersonation BFS fail-closes with a LOW evidence-gap instead of a falsely-clean `GRAPH_CLEAN` when any graph input is degraded (per-SA policy denied / custom-roles unavailable / list pagination-truncated); **(3) plugin 1024** a bucket whose DEFAULT object ACL is public (every future object born public) is now detected — and adds a structural read-only guarantee (a CI meta-test that fails the build on any mutating cloud call across all 28 plugins) plus the matching customer-facing read-only-credential requirement (EULA §5.5). No skill-logic change. **Plugin count UNCHANGED at 28; all six coverage matrices UNCHANGED** (SOC 2 + HIPAA + NIST CSF 2.0 + PCI DSS v4.0.1 + ISO 27001:2022 + CIS Controls v8).

package/SKILL.md

@@ -16,7 +16,7 @@ description: >
 
 # NSAuditor AI — Agent Skill
 
-> **Version:** 0.2.2 (post-EE 0.18.1 — GCP false-negative hardening II + read-only enforcement) · **Source:** [github.com/nsasoft/nsauditor-ai](https://github.com/nsasoft/nsauditor-ai) · **npm:** `nsauditor-ai` · **License:** MIT (CE)
+> **Version:** 0.2.3 (post-EE 0.18.2 — scan_cloud evidence-gap visibility end-to-end) · **Source:** [github.com/nsasoft/nsauditor-ai](https://github.com/nsasoft/nsauditor-ai) · **npm:** `nsauditor-ai` · **License:** MIT (CE)
 
 NSAuditor AI is a modular, AI-assisted network security audit platform with 27+ scanner
 plugins, CVE matching, MITRE ATT&CK mapping, and Zero Data Exfiltration by design. This

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "nsauditor-ai-agent-skill",
-  "version": "0.2.2",
+  "version": "0.2.3",
   "description": "AI Agent Skill for NSAuditor AI — gives any AI coding agent built-in knowledge of NSAuditor's MCP tools, schemas, plugins, and security audit workflows.",
   "keywords": [
     "nsauditor",