@blamejs/exceptd-skills 0.13.87 → 0.13.89

package/CHANGELOG.md

@@ -1,5 +1,13 @@
 # Changelog
 
+## 0.13.89 — 2026-05-25
+
+CVE catalog — NVIDIA Triton DALI backend memory safety. Completes the May 2026 Triton bulletin coverage with the three DALI (data-augmentation) backend flaws disclosed by researcher Navtej Kathuria, all fixed in r26.03: **CVE-2026-24213** (CWE-125 out-of-bounds read, NIST CVSS 9.8), **CVE-2026-24214** (CWE-190 integer overflow, NIST CVSS 9.8), and **CVE-2026-24215** (CWE-400 uncontrolled resource consumption / DoS, NIST CVSS 7.5). All process attacker-supplied inference input on a network-reachable backend. These are deliberate CVSS-versus-RWEP cases: NVD rates two of them CRITICAL, but with no CISA KEV listing, no confirmed in-the-wild exploitation, no public proof-of-concept, and a patch available, the Real-World Exploit Priority is P4 — the catalog scores priority on exploitation reality, not CVSS alone. Their shared zero-day lesson (NEW-CTRL-096) requires inference backends to bound and validate untrusted input size/shape and enforce resource limits, with the inference endpoint off untrusted networks. CVE count 358 → 361.
+
+## 0.13.88 — 2026-05-25
+
+CVE catalog — Hugging Face Transformers model-loader deserialization RCE. Adds the three ZDI-coordinated deserialization flaws in the foundational ML library's model loaders, all CWE-502 and fixed in 4.48.0: **CVE-2024-11392** (MobileViTV2 configuration files), **CVE-2024-11393** (MaskFormer model files), and **CVE-2024-11394** (Trax model files), each NIST CVSS 8.8 — loading a malicious model/config of the affected type executes attacker code in the user's process. All map MITRE ATLAS AML.T0010 / AML.T0011 / AML.T0011.000 and ATT&CK T1204 / T1059 / T1195.002, and they reuse the existing untrusted-model-artifact control (NEW-CTRL-091) — the same control that closes the Keras model-deserialization CVEs, because the class is "a model file is executable code", not a single loader. CVE count 355 → 358.
+
 ## 0.13.87 — 2026-05-25
 
 CVE catalog — Gradio file-access (Hugging Face Spaces secret theft). Adds the two Horizon3.ai-disclosed file-read flaws in Gradio, the ML demo/UI framework behind Hugging Face Spaces and countless public ML demos. **CVE-2024-1561** (CWE-22, NIST CVSS 7.5) — the `/component_server` endpoint invokes arbitrary Component methods with attacker-controlled arguments, abused via `move_resource_to_block_cache()` to read host files (and steal HF Spaces secrets); fixed in 4.13.0. **CVE-2023-51449** (CWE-22 + SSRF, NIST CVSS 7.5) — the `/file` route's directory-containment check was flawed, allowing arbitrary file read (and full-read SSRF) on a publicly reachable app; fixed in 4.11.0. Both map MITRE ATLAS AML.T0049 + AML.T0055 (unsecured credentials) and ATT&CK T1190 / T1083 / T1005; their shared zero-day lesson (NEW-CTRL-095) requires the framework's file-serving routes to enforce directory containment, not expose arbitrary method invocation or SSRF, and keep secret-bearing apps off untrusted networks. CVE count 353 → 355.

package/data/_indexes/_meta.json

@@ -1,21 +1,21 @@
 {
   "schema_version": "1.1.0",
-  "generated_at": "2026-05-25T20:20:21.658Z",
+  "generated_at": "2026-05-26T00:55:52.645Z",
   "generator": "scripts/build-indexes.js",
   "source_count": 54,
   "source_hashes": {
-    "manifest.json": "439dedebe750531eb56ca6945a3f21486234d95f6517ac710d111980f358a079",
-    "data/atlas-ttps.json": "7e81fb7e2202749c3fc6e68fa437d1f6d79ba3b9a7cba999127143205f366041",
-    "data/attack-techniques.json": "490aed27f8f62cbf5f077ff66eaa095598097952fa2a332d0f25027b07c51e1a",
-    "data/cve-catalog.json": "31f12e733ac9f24650c62997f72fa476d88ec3cec8fbaad60e43b2cbd9ad3cc9",
-    "data/cwe-catalog.json": "219ddfa7e4a2464d9e217fb19f181cf6594edb369604780cbb226bbb31b8f439",
+    "manifest.json": "4696ac730af56b3d1e3792f13011dbe9df415fa6ed6db039f3d5831f98dfd4d7",
+    "data/atlas-ttps.json": "35db40b134ee32ed429a62970c2b05c8937cd8c3df64f5f84e5d725df7cbb5a5",
+    "data/attack-techniques.json": "fef078610c1533ffc17da09a0720ad86cd1a86baadecac9776501059d9ab99b9",
+    "data/cve-catalog.json": "6574013902f68e32cb3b495eef64ca62a7441ef1b0f904a14065a5d08dfa96a5",
+    "data/cwe-catalog.json": "253af97e89506cf61de0e455f1d42a25c8b04594b0e8795737cbe9dc696d09b1",
     "data/d3fend-catalog.json": "9a54bccb9f24f84b32024216cc3f53819a053721ac8ab43c326859e68fc0ffaf",
     "data/dlp-controls.json": "d2406c482dddd30e49203879999dc4b3a7fd4d0494d6a61d86b91ee76415df19",
     "data/exploit-availability.json": "ec2656f0d9a893610e27b43eb6035fe9b18e057c9f6dfaac7e7d4959bbcbb795",
-    "data/framework-control-gaps.json": "1194be6f0806b027ed2d6dbdf106e291727ed82da480badb49d3b2779ed70951",
+    "data/framework-control-gaps.json": "9cf7fce07276c1049f7cd28e20becedf981a18709f094cae9686d452031213e5",
     "data/global-frameworks.json": "9ba563a85f7f8d6c3c957de64945e20925a89d0ed6ea6fc561cf093811acf558",
     "data/rfc-references.json": "66ef2e1f444a2cf0c2700a754f0a66030bb8a91d9e68394b9537ea1fe8b904fe",
-    "data/zeroday-lessons.json": "885bd899b2475cab39d4c1f1a1bf8618750f5492e58d616cd7048e5cb715c294",
+    "data/zeroday-lessons.json": "a153d695e8f088e1da3c6fbc63285f5eff330fca192df2f7a9a5eac70abf4017",
     "skills/kernel-lpe-triage/skill.md": "08b3e9815ba481c57c80f5fc0ccbf5bb7cbb41f570c235ba6ff9596b8c07354d",
     "skills/ai-attack-surface/skill.md": "c4c1eb22a38ca7a959b5725222bab8fbd4f4044a548a93f3e288e6f698334b72",
     "skills/mcp-agent-trust/skill.md": "89ac89084391d2341b6513fefb1be2d36b93de1c130f057696219c1c59440f13",
@@ -72,7 +72,7 @@
       "dlp_refs": 0
     },
     "trigger_table_entries": 538,
-    "chains_cve_entries": 344,
+    "chains_cve_entries": 350,
     "chains_cwe_entries": 171,
     "jurisdictions_indexed": 29,
     "handoff_dag_nodes": 42,

package/data/_indexes/activity-feed.json

@@ -149,7 +149,7 @@
       "artifact": "data/cve-catalog.json",
       "path": "data/cve-catalog.json",
       "schema_version": "1.0.0",
-      "entry_count": 355
+      "entry_count": 361
     },
     {
       "date": "2026-05-18",
@@ -165,7 +165,7 @@
       "artifact": "data/zeroday-lessons.json",
       "path": "data/zeroday-lessons.json",
       "schema_version": "1.1.0",
-      "entry_count": 350
+      "entry_count": 356
     },
     {
       "date": "2026-05-17",

package/data/_indexes/catalog-summaries.json

@@ -62,7 +62,7 @@
         "rebuild_after_days": 365,
         "note": "Per-entry last_verified governs decay. Skills depending on this catalog must check entry freshness before high-stakes use."
       },
-      "entry_count": 355,
+      "entry_count": 361,
       "sample_keys": [
         "CVE-2025-53773",
         "CVE-2026-30615",
@@ -238,7 +238,7 @@
         "rebuild_after_days": 365,
         "note": "Per-entry last_verified governs decay. Skills depending on this catalog must check entry freshness before high-stakes use."
       },
-      "entry_count": 350,
+      "entry_count": 356,
       "sample_keys": [
         "CVE-2026-31431",
         "CVE-2025-53773",