open/ms-ai-architect
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity Actions

chore(release): v1.16.4 — KB-refresh MEDIUM-tier (33 filer): prompt-caching/Realtime/Guardrails/PTU/Foundry-deprecations + 18 TRYGT (medium 33→0)

Browse source
This commit is contained in:
Kjell Tore Guttormsen 2026-06-24 11:42:34 +02:00
commit 491f2b85cd
3 changed files with 18 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

2
.claude-plugin/plugin.json
View file

@ -1,6 +1,6 @@
{
"name": "ms-ai-architect",
"version": "1.16.3",
"version": "1.16.4",
"description": "Microsoft AI Solution Architect - structured architecture guidance for the full Microsoft AI stack",
"author": {
"name": "Kjell Tore Guttormsen"

15
CHANGELOG.md
View file

@ -5,6 +5,21 @@ All notable changes to this project will be documented in this file.
The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
## [1.16.4] - 2026-06-24
KB-refresh-patch. MEDIUM-tier (33 reference docs) verifisert mot Microsoft Learn (post-1.16.3 baseline). Ingen kommando-, agent- eller skill-endringer — kun reference docs. 18 filer var TRYGT (innhold gjeldende, kun header-bump); 15 filer hadde reell drift. Hver `VERIFY`-påstand fra de 7 parallelle drift-rapportene ble re-bekreftet av main-agent mot sitert Microsoft Learn-kilde FØR skriving; disclaimed/illustrative priser urørt; GPT-4.1-pricing-påstand forble uverifiserbar (kun community-kilder) og ble IKKE endret; alle endringer review-gated på `git diff` + plugin-validering (239/0). Medium-count 33 → 0 (needing-update 85 → 52).
### Changed
- **Prompt-caching:** `token-optimization` — retention-modell omskrevet: in-memory (5–10 min, ≤1t, default t.o.m. gpt-5.4) vs **extended retention 24t** (`prompt_cache_retention`, default for nyere modeller); in-memory støttes av «GPT-4o eller nyere».
- **Realtime:** `real-time-reasoning` — la til `gpt-realtime-1.5` (2026-02-23); Realtime API flyttet til **GA-endepunkt** `/openai/v1` (ikke lenger date-baserte api-version); preview→GA-statusnyanse.
- **Guardrails-rebranding:** `regulatory-and-compliance` + `azure-ai-foundry` — «content filters» → **Guardrails (and controls)**, default `Microsoft.DefaultV2`, fire intervensjonspunkter; data-privacy-kildetittel → «Models sold by Azure».
- **GPU/PTU:** `gpu-compute-sizing` — `ND96asr_v4` = 8×A100 **40GB = 320 GB** (var 640 GB); `load-testing` — PTU min deployment Global/Data Zone **15** (increment 5), 50 kun Regional.
- **Foundry-deprecations:** `azure-ai-foundry` — **Deep Research-tool** og **Connected agents (classic)** deprecated (retire 2027-03-31), **Prompt Flow**-retirement 2027-04-20, **A2A-tool nå i Java SDK**.
- **Agent-evaluering/-memory:** `agent-evaluation` — nye agent-evaluatorer (Task Completion/Customer Satisfaction/Tool Selection m.fl.) + `FoundryEvals` + `gpt-4.1-mini` judge; `agent-memory` — **procedural memory** (3. type), TTL/CRUD, **Norway East**-region bekreftet (VNet ikke støttet).
- **Øvrig faktafiks:** `multi-turn` RPM/TPM-forhold varierer per modell; `structured-output` usupporterte keywords utvidet + `gpt-4o-mini-audio-preview`; `temperature-sampling` Copilot Studio prompt builder-temp + fjernet uoffisielt M365-tall; `rag-hallucination` groundedness `mitigating`/`correctionText`; `model-versioning` MLflow stages legacy i UC; `speaker-recognition` **Limited Access**-status; `chain-of-thought` reasoning_effort per-modell-nyanse.
- **Header-bump (TRYGT, 18 filer):** domain-specific, few-shot, function-calling, grounding, role-playing, system-message-design, async-processing, batch-api-usage, concurrent-request, performance-benchmarking, response-chunking, throughput-optimization, token-per-second, late-chunking, rag-evaluation, self-reflective-rag, streaming-rag, feedback-loops.
## [1.16.3] - 2026-06-24
KB-refresh-patch. HIGH-tier (11 reference docs) + 1 dateless kjernefil (`ms-ai-advisor/references/architecture/security.md`), flagget av endrings-detektoren mot Microsoft Learn-sitemaps (post-1.16.2 baseline), er verifisert og oppdatert. Ingen kommando-, agent- eller skill-endringer — kun reference docs. Hver `VERIFY`-påstand fra de parallelle drift-rapportene ble bekreftet mot sitert Microsoft Learn-kilde FØR skriving (main-agent som verifiseringsgate); disclaimed/illustrative priser urørt; alle endringer review-gated på `git diff` + plugin-validering (239/0). High-count 11 → 0, Unverified 1 → 0 (needing-update 97 → 85).

3
README.md
View file

@ -6,7 +6,7 @@
*AI-generated: all code produced by Claude Code through dialog-driven development. [Full disclosure →](../../README.md#ai-generated-code-disclosure)*
![Version](https://img.shields.io/badge/version-1.16.3-blue)
![Version](https://img.shields.io/badge/version-1.16.4-blue)
![Platform](https://img.shields.io/badge/platform-Claude_Code_Plugin-purple)
![Docs](https://img.shields.io/badge/reference_docs-389-green)
![Agents](https://img.shields.io/badge/agents-12-orange)
@ -659,6 +659,7 @@ Category-to-skill routing is defined in `scripts/kb-update/data/domain-taxonomy.
| Version | Date | Highlights |
|---------|------|-----------|
| **1.16.4** | 2026-06-24 | KB-refresh-patch (kun reference docs). **MEDIUM-tier 33 filer** verifisert mot Microsoft Learn (18 TRYGT header-bump, 15 reell drift). **Prompt-caching** in-memory vs extended 24t (`prompt_cache_retention`); **Realtime** `gpt-realtime-1.5` + GA-endepunkt `/openai/v1`; **Guardrails**-rebranding (tidl. content filters, `Microsoft.DefaultV2`); `ND96asr_v4` 320 GB (var 640); PTU min Global/Data Zone 15 (var 50); **Deep Research + Connected agents deprecated**, Prompt Flow-retirement 2027-04-20, A2A Java; nye agent-evaluatorer + `FoundryEvals`; **procedural memory** + Norway East; **Speaker Recognition Limited Access**. GPT-4.1-pricing forble uverifiserbar (community-kilder) → urørt. Hver `VERIFY` re-bekreftet av main-agent mot kilde FØR skriving; `git diff` + validering 239/0. Medium 33 → 0 (needing-update 85 → 52). |
| **1.16.3** | 2026-06-24 | KB-refresh-patch (kun reference docs). **HIGH-tier 11 filer + 1 dateless kjernefil** verifisert mot Microsoft Learn. **Faktafiks:** Content Safety image-modeller Medium-terskel (ikke «Low»); `gpt-4.1-mini` = GPT-4.1-variant (ikke o-series reasoning); CMK Key Vault Azure RBAC «Key Vault Crypto Service Encryption User» som primæralternativ (ikke «legacy access policies»). **Lagt til:** Task Adherence (preview) i guardrails-risk-kategorier, Prompt Flow retirement 2027-04-20, 5. data-privacy-garanti, «Models sold by Azure»-terminologi, `security.md` maskinlesbar header. **Avkreftet ved verifisering:** Content Safety Analyze Text 10K-tegn-grensen var korrekt (drift-rapportens «1000» = feillesing av Transparency Note-anbefaling). Hver `VERIFY` bekreftet mot kilde FØR skriving; disclaimed priser urørt; `git diff` + plugin-validering 239/0. High 11 → 0, Unverified 1 → 0 (needing-update 97 → 85). |
| **1.16.2** | 2026-06-24 | KB-refresh-patch (kun reference docs — ingen kommando-/agent-/skill-endringer). **Critical cost-klyngen** (8 filer i `cost-optimization/`) verifisert + oppdatert mot Microsoft Learn: Spillover GA, **Priority processing** (4. deployment-type), **Data Zone Batch** (SKU `DataZoneBatch`, EU/US-datasone), modell-lineup GPT-5 → 5.1/5.5, extended prompt-cache 24t, RFT-kostnadsmodell ($5000 per-job-tak), batch fail-fast-retry (server fail-fast + klient-side backoff), On Your Data-pensjon (2026-10-14). **Kalibrering:** fjernet uverifiserbart «175B+» (GPT-4o), «GPT-4 utfaset» → «legacy», og avkreftet én drift-påstand ved verifisering (Phi-3/Phi-2/Falcon-7B fortsatt i Foundry-katalogen → ingen «retired»-merking). Hver `VERIFY`-påstand bekreftet mot kilde FØR skriving; disclaimed priser urørt; `git diff` + plugin-validering 239/0. Critical-count 8 → 0. |
| **1.16.1** | 2026-06-23 | KB-refresh-patch (kun reference docs — ingen kommando-/agent-/skill-endringer). **Modellkatalog** GPT-5.1 → GPT-5.5 i 3 filer (`batch-processing-cost-reduction.md`, `model-selection-price-performance.md`, `azure-ai-foundry.md`). **Foundry-navnesveip** «Azure AI Foundry» → «Microsoft Foundry» på tvers av 233 reference docs (810 forekomster). **RBAC** «Azure AI User» → «Foundry User» (`foundry-workflows-visual-orchestration.md`; cost-governance-mappingen bevart). Hvert premiss verifisert mot Microsoft Learn før sveip; `git diff` + plugin-validering + deterministisk re-score (skills uendret, alle ≥90 %). |