nsauditor-ai-agent-skill 0.1.63 → 0.1.65

package/CHANGELOG.md

@@ -4,6 +4,12 @@ Release notes for **`nsauditor-ai-agent-skill`** — installable knowledge packa
 
 ---
 
+## 0.1.65 (2026-05-31) — Paired-release pin for EE 0.16.6 + CE 0.1.97 — EE false-clean regression hotfix + two over-reporting folds (CloudTrail soft budget = 0.8×min(PLUGIN_TIMEOUT_MS, CLOUD_PLUGIN_TIMEOUT_MS) so it stays below the manager wall on BOTH the CLI and MCP scan paths — closes the 0.16.5 hard-cancel regression the published-build re-smoke caught; plugin 1110 no longer flags KMS grant-decrypt Pacu P-16 stealth paths on AWS-managed CMKs; compliance-engine PASS-tier substrate-evidence no longer counted as a violation). Internal EE engine/plugin change; no tool surface or agent-behavior change. Plugin count UNCHANGED (28); all six matrices UNCHANGED. Full regression 6808/6812 GREEN (zero new failures, zero verdict shifts). SKILL.md/references unchanged.
+
+## 0.1.64 (2026-05-31) — Paired-release pin for EE 0.16.5 + CE 0.1.96 — EE compliance-mapping false-clean fixes (perimeter exposures route cross-framework to PCI/ISO/NIST; CloudTrail fails-closed on abort; GCP Cloud Storage emits an evidence-gap on enumeration failure; internal review-process markers + repo paths scrubbed from the shipped framework JSONs). Internal EE engine/JSON change; no tool surface or agent-behavior change. Plugin count UNCHANGED (28); all six matrices UNCHANGED. SKILL.md/references unchanged.
+
+## 0.1.63 (2026-05-30) — Paired-release pin for EE 0.16.4 + CE 0.1.95 — MCP `scan_cloud` now surfaces cloud findings reliably (CE-led false-clean fix: builds `findingsSummary` directly from the scan results instead of the network-host concluder, which silently dropped cloud findings). Internal CE engine change; no tool surface or behavior change for the agent. Plugin count UNCHANGED (28); all six matrices UNCHANGED. SKILL.md/references unchanged.
+
 ## 0.1.62 (2026-05-30) — Paired-release pin for EE 0.16.3 + CE 0.1.94 — MCP `scan_cloud` now runs its cloud plugins concurrently with a per-plugin timeout (`CLOUD_SCAN_CONCURRENCY` / `CLOUD_PLUGIN_TIMEOUT_MS`), so a full AWS/GCP/Azure account audit finishes within Claude Desktop's ~60s tool-call limit. Internal CE engine change; no tool surface or behavior change for the agent. Plugin count UNCHANGED (28); all six matrices UNCHANGED. SKILL.md/references unchanged.
 
 ## 0.1.61 (2026-05-30) — **HOTFIX: SKILL.md `description` trimmed to fit the 1024-char skill-upload limit.** The frontmatter `description` had grown to 1241 chars, so Claude Desktop rejected the skill upload (`field 'description' in SKILL.md must be at most 1024 characters`). Rewrote it to 967 chars while keeping the high-value trigger keywords and adding the new cloud-audit triggers (`scan_cloud`, "audit my AWS/GCP/Azure account", "cloud compliance"). No body/content change vs 0.1.60 (still teaches `scan_cloud`); pairs the same EE 0.16.2 + CE 0.1.93.

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "nsauditor-ai-agent-skill",
-  "version": "0.1.63",
+  "version": "0.1.65",
   "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",