fix(ms-ai-architect): R11 — første korpus-edit; 4 ratifiserte O2-subtraksjoner anvendt (idx 17, 19, 33, 14-redusert)

Operatør-ratifisert 2026-08-03. Sletter udokumenterte påstander fra tre
KB-filer: score-threshold-båndene og 'Automatic indexing av vectors'
(rag-caching-optimization), to ikke-støttede AISPM-attribusjoner
(ai-threat-modeling-stride), og SharePoint som feedback-lagring
(feedback-loops). idx 14 er den REDUSERTE subtraksjonen — 'Automatically'
beholdes, siden linje 566 hevder automatikk.

Ikke ratifisert og ikke anvendt: idx 26, 27, 36, 18.

Driver ankrer på file_text_verbatim, aldri linjenummer, og avbryter uten
å skrive ved tvetydig anker, ikke-sletting eller ny ordform. 11 tester.
Suite 1032/1032. [skip-docs]

skills/ms-ai-engineering/references/mlops-genaiops/feedback-loops-continuous-improvement.md

@@ -552,7 +552,7 @@ ml_client.schedules.begin_create_or_update(model_monitor)
 | Komponent | Power Platform-tjeneste | Formål |
 |-----------|-------------------------|--------|
 | **Automated feedback collection** | Power Automate | Route low-confidence predictions til human review |
-| **Storage** | Dataverse / SharePoint | Lagre feedback data |
+| **Storage** | Dataverse | Lagre feedback data |
 | **Model improvement** | AI Builder Feedback Loop | Automatically add reviewed samples to training set |
 | **Retraining** | AI Builder | Manual/scheduled retraining |
 

skills/ms-ai-engineering/references/rag-architecture/rag-caching-optimization.md

@@ -251,9 +251,6 @@ Outbound (cache store):
 ```
 
 **Score Threshold Tuning** (APIM `score-threshold` er en DISTANSE: lavere = strengere, krever høyere semantisk likhet):
-- 0.1-0.2 → Strict matching, lavere hit rate, høy relevance
-- 0.3-0.5 → Balanced, medium hit rate, god relevance
-- 0.6-0.8 → Liberal matching, høyere hit rate, noe lavere relevance
 
 **Verified** (Microsoft Learn - Enable semantic caching for LLM APIs)
 
@@ -294,7 +291,6 @@ def query_cache(prompt_vector, similarity_threshold=0.15, top_k=5):
 
 **Fordeler:**
 - Globally distributed, multi-region writes
-- Automatic indexing av vectors
 - 99.999% SLA med multi-region setup
 - Built-in TTL support
 

skills/ms-ai-security/references/ai-security-engineering/ai-threat-modeling-stride.md

@@ -208,10 +208,9 @@ STRIDE Mapping: Tampering
 ### Microsoft Defender for Cloud — AI Security Posture Management
 
 **Capabilities:** *(Verified MCP 2026-04)*
-- Automated detection of AI workloads across Azure subscriptions (via Azure Resource Graph)
+- Automated detection of AI workloads across Azure subscriptions
 - AI security posture management: automate detection and remediation of generative AI risks
 - Security recommendations for AI models, data stores, network isolation
-- Integration with Purview for data classification, DLP og Insider Risk Management for prompt-based data exfiltration
 
 **Threat Modeling Integration:**
 ```plaintext