chore(ms-ai-architect): sanitize all private references for open-source
README.md: badges updated (1.7.0/387/12), installation URL updated to ktg-plugin-marketplace, added ai-act-assessor to agent table, updated skill ref counts, updated hooks section, updated category-skill-map path. CLAUDE.md: fix agent model column (sonnet->opus), remove Linear section, fix manual test path to generic placeholder. commands/generate-skills.md: orchestrator paths updated to scripts/skill-gen. commands/export.md: add Bash scope guardrail (security scan finding). docs: replace GitHub and ktg-privat URLs with Forgejo, replace personal paths. scripts/skill-gen/manifest.json: rename ktg-privat ID. skills: remove Linear tagging reference, add supply chain warnings. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
Kjell Tore Guttormsen
parent
ed72d633e8
commit
bd22b29a21
10 changed files with 55 additions and 58 deletions
|
|
@ -48,18 +48,18 @@ Tilbyr strukturert arkitekturveiledning for Microsoft AI-stakken:
|
|||
|
||||
| Agent | Formål | Modell |
|
||||
|-------|--------|--------|
|
||||
| `research-agent` | MCP-isolert research med microsoft-learn | sonnet |
|
||||
| `security-assessment-agent` | 6-dimensjons sikkerhetsrammeverk med 1-5 scoring | sonnet |
|
||||
| `cost-estimation-agent` | Kostnadsestimering i NOK med TCO-sammenligning | sonnet |
|
||||
| `adr-writer-agent` | ADR-generering i MADR v3.0-format | sonnet |
|
||||
| `license-mapper-agent` | Kryssreferering av lisenser mot plattformkapabiliteter | sonnet |
|
||||
| `diagram-generation-agent` | Arkitekturdiagrammer med Imagen 3 via mcp-image | sonnet |
|
||||
| `architecture-review-agent` | Arkitekturgjennomgang mot Digdir, AI Act, NSM, Schrems II | sonnet |
|
||||
| `ros-analysis-agent` | ROS-analyse med 7 dimensjoner, NS 5814-metodikk og AI-trusselbibliotek | sonnet |
|
||||
| `dpia-agent` | DPIA/PVK for AI-systemer med risikomatrise og tiltakstabell | sonnet |
|
||||
| `summary-agent` | Teknisk sammendrag og beslutningsnotat fra arkitekturvurderinger | sonnet |
|
||||
| `ai-act-assessor` | EU AI Act-klassifisering, forpliktelser og compliance-vurdering | sonnet |
|
||||
| `onboarding-agent` | Strukturert onboarding-intervju for virksomhetstilpasning | sonnet |
|
||||
| `research-agent` | MCP-isolert research med microsoft-learn | opus |
|
||||
| `security-assessment-agent` | 6-dimensjons sikkerhetsrammeverk med 1-5 scoring | opus |
|
||||
| `cost-estimation-agent` | Kostnadsestimering i NOK med TCO-sammenligning | opus |
|
||||
| `adr-writer-agent` | ADR-generering i MADR v3.0-format | opus |
|
||||
| `license-mapper-agent` | Kryssreferering av lisenser mot plattformkapabiliteter | opus |
|
||||
| `diagram-generation-agent` | Arkitekturdiagrammer med Imagen 3 via mcp-image | opus |
|
||||
| `architecture-review-agent` | Arkitekturgjennomgang mot Digdir, AI Act, NSM, Schrems II | opus |
|
||||
| `ros-analysis-agent` | ROS-analyse med 7 dimensjoner, NS 5814-metodikk og AI-trusselbibliotek | opus |
|
||||
| `dpia-agent` | DPIA/PVK for AI-systemer med risikomatrise og tiltakstabell | opus |
|
||||
| `summary-agent` | Teknisk sammendrag og beslutningsnotat fra arkitekturvurderinger | opus |
|
||||
| `ai-act-assessor` | EU AI Act-klassifisering, forpliktelser og compliance-vurdering | opus |
|
||||
| `onboarding-agent` | Strukturert onboarding-intervju for virksomhetstilpasning | opus |
|
||||
|
||||
## Skills (5 domenespesifikke)
|
||||
|
||||
|
|
@ -147,7 +147,7 @@ Fixture-basert validering av agent-output (sikkerhet, kostnad, sammendrag). Test
|
|||
#### Manuell test
|
||||
```bash
|
||||
# Test at plugin registreres
|
||||
cd /Users/ktg/repos/plugins/ktg-privat
|
||||
cd <plugin-root>
|
||||
claude --plugin ./plugins/ms-ai-architect
|
||||
|
||||
# Kjør hovedcommand
|
||||
|
|
@ -198,7 +198,3 @@ Interaktiv 5-stegs arkitektur-pipeline for Azure AI-beslutninger.
|
|||
|
||||
**Tilsynsmyndigheter:** Datatilsynet (personvern), nasjonal AI-tilsynsmyndighet (under etablering), sektortilsyn.
|
||||
|
||||
## Linear
|
||||
|
||||
Project: MS AI Architect Plugin Suite
|
||||
Issues tagges med `🏛️ ARCHITECT` label.
|
||||
|
|
|
|||
|
|
@ -2,10 +2,10 @@
|
|||
|
||||
> Your virtual Microsoft AI solution architect — meet **Cosmo Skyberg**.
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
A Claude Code plugin that provides structured architecture guidance across the full Microsoft AI stack. Cosmo Skyberg is a methodical, opinionated architect persona who understands the problem before recommending technology, verifies claims against live Microsoft Learn documentation via MCP, and delivers assessments calibrated for Norwegian public sector governance — while remaining useful for any enterprise context.
|
||||
|
|
@ -65,7 +65,7 @@ Key capabilities:
|
|||
### Installation
|
||||
|
||||
```bash
|
||||
claude plugin add ktg-privat/ms-ai-architect
|
||||
claude plugin marketplace add https://git.fromaitochitta.com/open/ktg-plugin-marketplace.git
|
||||
```
|
||||
|
||||
Or add to your `~/.claude/settings.json`:
|
||||
|
|
@ -73,7 +73,7 @@ Or add to your `~/.claude/settings.json`:
|
|||
```json
|
||||
{
|
||||
"enabledPlugins": {
|
||||
"ms-ai-architect@ktg-privat": true
|
||||
"ms-ai-architect@ktg-plugin-marketplace": true
|
||||
}
|
||||
}
|
||||
```
|
||||
|
|
@ -146,7 +146,7 @@ Cosmo will ask clarifying questions about your business need, licenses, data sou
|
|||
|
||||
## Agent Architecture
|
||||
|
||||
The plugin delegates specialized work to 11 purpose-built agents. Each agent has its own knowledge base routing, model assignment, and tool access.
|
||||
The plugin delegates specialized work to 12 purpose-built agents. Each agent has its own knowledge base routing, model assignment, and tool access.
|
||||
|
||||
| Agent | Role | KB Sources | Triggered By |
|
||||
|-------|------|------------|--------------|
|
||||
|
|
@ -161,6 +161,7 @@ The plugin delegates specialized work to 11 purpose-built agents. Each agent has
|
|||
| `diagram-generation-agent` | Architecture diagrams via Imagen 3 / Mermaid | Prompt templates | `/architect:diagram` |
|
||||
| `summary-agent` | Executive summary and decision memo synthesis | All assessment outputs (incl. ROS) | `/architect:summary` |
|
||||
| `onboarding-agent` | 5-phase structured org interview | Writes org/*.md | `/architect:onboard` |
|
||||
| `ai-act-assessor` | EU AI Act classification, obligations, and compliance assessment | ms-ai-governance (ai-act-*) | `/architect:classify`, `/architect:requirements`, `/architect:transparency`, `/architect:frimpact`, `/architect:conformity` |
|
||||
|
||||
### Orchestration Pattern
|
||||
|
||||
|
|
@ -193,25 +194,25 @@ The orchestrator creates a `.work/` directory for intermediate results, delegate
|
|||
|
||||
## Knowledge Base
|
||||
|
||||
The plugin includes **380 reference documents** organized across 5 domain-specific skills:
|
||||
The plugin includes **387 reference documents** organized across 5 domain-specific skills:
|
||||
|
||||
| Skill | Domain | Refs | User Intent |
|
||||
|-------|--------|------|-------------|
|
||||
| `ms-ai-advisor` | Cosmo persona, 7-phase workflow, platform selection | 62 | "Help me choose" |
|
||||
| `ms-ai-engineering` | RAG, agents, Azure AI Services, data, MLOps, multimodal | 153 | "How do I build this?" |
|
||||
| `ms-ai-governance` | Norwegian public sector governance, EU regulations, responsible AI, ROS | 71 | "Is this legal/safe?" |
|
||||
| `ms-ai-governance` | Norwegian public sector governance, EU regulations, responsible AI, ROS | 78 | "Is this legal/safe?" |
|
||||
| `ms-ai-security` | Security scoring (6×5), cost estimation (P10/P50/P90) | 60 | "Is this safe?" |
|
||||
| `ms-ai-infrastructure` | BCDR, hybrid/edge, sovereign cloud | 34 | "How do I operate this?" |
|
||||
|
||||
### ms-ai-advisor (61 refs)
|
||||
### ms-ai-advisor (62 refs)
|
||||
|
||||
Architecture decision trees, platform comparison matrices, Cosmo persona definition, cost models, migration patterns.
|
||||
|
||||
### ms-ai-engineering (149 refs)
|
||||
### ms-ai-engineering (153 refs)
|
||||
|
||||
RAG implementation patterns, agent orchestration, Azure AI Foundry, Copilot Studio extensibility, AI Builder, multimodal processing, Semantic Kernel, MLOps pipelines.
|
||||
|
||||
### ms-ai-governance (71 refs)
|
||||
### ms-ai-governance (78 refs)
|
||||
|
||||
Norwegian public sector governance (Digdir, DFØ), EU AI Act (Annex III checklist), responsible AI frameworks, GDPR/Schrems II compliance, Utredningsinstruksen alignment. Includes a comprehensive **ROS analysis framework** with 7 new reference documents: AI threat library (49 threats across 7 categories), NS 5814/ISO 31000 methodology guide, 7×5 scoring rubrics, sector-specific checklists (health, transport, finance, justice, education), report templates, DPIA/security integration patterns, and MAESTRO multi-agent security model.
|
||||
|
||||
|
|
@ -329,25 +330,15 @@ These MCP servers enhance the plugin's capabilities but are not required:
|
|||
|
||||
## Hooks & Safety
|
||||
|
||||
Three runtime hooks provide session context and safety guardrails:
|
||||
Two runtime hooks provide session context and safety guardrails:
|
||||
|
||||
| Event | Script | Purpose | Behavior |
|
||||
|-------|--------|---------|----------|
|
||||
| `SessionStart` | `session-start-context.mjs` | Show active investigations + KB freshness | Advisory — displays context |
|
||||
| `PreToolUse` (Edit\|Write) | `pre-edit-secrets.mjs` | Block Azure keys, tokens, credentials from being written | **Blocking** — prevents write |
|
||||
| `Stop` | `stop-assessment-reminder.mjs` | Remind about uncommitted assessments and next steps | Advisory — displays reminder |
|
||||
|
||||
### Secrets Detection
|
||||
|
||||
The `pre-edit-secrets` hook scans all Edit and Write operations for patterns matching:
|
||||
|
||||
- Azure subscription keys and connection strings
|
||||
- Bearer tokens and API keys
|
||||
- Service principal credentials
|
||||
- SAS tokens and storage account keys
|
||||
|
||||
> [!IMPORTANT]
|
||||
> The secrets hook is **blocking** — it will prevent the write operation if a secret pattern is detected. This is a safety net, not a replacement for proper secrets management with `.env` files.
|
||||
> [!TIP]
|
||||
> For secrets scanning across all plugins, use the [llm-security plugin](https://git.fromaitochitta.com/open/ktg-plugin-marketplace) which provides byte-level secrets detection as a blocking PreToolUse hook.
|
||||
|
||||
---
|
||||
|
||||
|
|
@ -519,7 +510,7 @@ bash scripts/kb-staleness-check.sh --json --output report.json
|
|||
/architect:generate-skills --update
|
||||
```
|
||||
|
||||
Category-to-skill routing is defined in `category-skill-map.json` (20 categories mapped to 5 skills), used by the generate-skills workflow to place new reference documents in the correct skill directory.
|
||||
Category-to-skill routing is defined in `scripts/skill-gen/category-skill-map.json` (20 categories mapped to 5 skills), used by the generate-skills workflow to place new reference documents in the correct skill directory.
|
||||
|
||||
---
|
||||
|
||||
|
|
|
|||
|
|
@ -10,6 +10,8 @@ model: opus
|
|||
|
||||
Eksporter et markdown-dokument til profesjonell PDF med A4-layout, tabellformatering og fargekodet scoring.
|
||||
|
||||
> **Bash scope:** Bash brukes utelukkende til å kjøre `scripts/export-pdf.py`. Ingen andre shell-operasjoner skal utføres.
|
||||
|
||||
## Språk og encoding
|
||||
|
||||
**VIKTIG:** Bruk norske tegn (æ, ø, å) korrekt i all output.
|
||||
|
|
|
|||
|
|
@ -20,13 +20,13 @@ Dette gir ~15-20 skills per sesjon istedenfor ~5.
|
|||
|
||||
## Oppstart
|
||||
|
||||
1. **Les state:** `orchestrator/skill-gen/state.json` — hva er allerede generert?
|
||||
2. **Les manifest:** `orchestrator/skill-gen/manifest.json` — hvilke skills finnes?
|
||||
1. **Les state:** `scripts/skill-gen/state.json` — hva er allerede generert?
|
||||
2. **Les manifest:** `scripts/skill-gen/manifest.json` — hvilke skills finnes?
|
||||
3. **Sjekk om manifest er komplett:**
|
||||
- Les `orchestrator/skill-gen/categories.json` for å se alle 15 kategorier
|
||||
- Les `scripts/skill-gen/categories.json` for å se alle 15 kategorier
|
||||
- Hvis en kategori mangler i manifest, ekspander den ved å kjøre:
|
||||
```bash
|
||||
./orchestrator/skill-gen/expand-categories.sh <category-key>
|
||||
./scripts/skill-gen/expand-categories.sh <category-key>
|
||||
```
|
||||
- Gjenta for alle manglende kategorier
|
||||
4. **Beregn pending:** Alle skills i manifest som IKKE er i state.completed
|
||||
|
|
@ -182,7 +182,7 @@ Task(general-purpose, sonnet): "Research + write skill: Multi-Index..."
|
|||
|
||||
2. **Commit:**
|
||||
```bash
|
||||
git add skills/ms-ai-*/references/<dirs>/ orchestrator/skill-gen/state.json
|
||||
git add skills/ms-ai-*/references/<dirs>/ scripts/skill-gen/state.json
|
||||
git commit -m "docs(architect): generate N knowledge skills (category-names)"
|
||||
```
|
||||
|
||||
|
|
|
|||
|
|
@ -455,7 +455,7 @@ bash tests/test-ai-act-output.sh
|
|||
### STEG 9: Kjør validate-plugin.sh
|
||||
|
||||
```bash
|
||||
cd /Users/ktg/.claude/plugins/marketplaces/ktg-privat/plugins/ms-ai-architect
|
||||
cd <plugin-root>
|
||||
bash tests/validate-plugin.sh
|
||||
```
|
||||
|
||||
|
|
|
|||
|
|
@ -16,7 +16,7 @@ Open PowerShell:
|
|||
New-Item -ItemType Directory -Force -Path "$env:USERPROFILE\.claude\plugins\marketplaces"
|
||||
|
||||
# Clone
|
||||
git clone https://github.com/guttormsen108/ktg-privat.git "$env:USERPROFILE\.claude\plugins\marketplaces\ktg-privat"
|
||||
git clone https://git.fromaitochitta.com/open/ktg-plugin-marketplace.git "$env:USERPROFILE\.claude\plugins\marketplaces\ktg-plugin-marketplace"
|
||||
```
|
||||
|
||||
Edit `%USERPROFILE%\.claude\settings.json` (create if it doesn't exist):
|
||||
|
|
@ -24,7 +24,7 @@ Edit `%USERPROFILE%\.claude\settings.json` (create if it doesn't exist):
|
|||
```json
|
||||
{
|
||||
"enabledPlugins": {
|
||||
"ms-ai-architect@ktg-privat": true
|
||||
"ms-ai-architect@ktg-plugin-marketplace": true
|
||||
},
|
||||
"mcpServers": {
|
||||
"microsoft-learn": {
|
||||
|
|
@ -40,7 +40,7 @@ Edit `%USERPROFILE%\.claude\settings.json` (create if it doesn't exist):
|
|||
## Step 2: Verify
|
||||
|
||||
```powershell
|
||||
cd "$env:USERPROFILE\.claude\plugins\marketplaces\ktg-privat"
|
||||
cd "$env:USERPROFILE\.claude\plugins\marketplaces\ktg-plugin-marketplace"
|
||||
claude
|
||||
```
|
||||
|
||||
|
|
@ -103,7 +103,7 @@ After implementation, run validation in **Git Bash** (not PowerShell — the tes
|
|||
|
||||
```bash
|
||||
# Open Git Bash from Start menu, then:
|
||||
cd ~/.claude/plugins/marketplaces/ktg-privat
|
||||
cd ~/.claude/plugins/marketplaces/ktg-plugin-marketplace
|
||||
|
||||
# Plugin structure validation
|
||||
bash plugins/ms-ai-architect/tests/validate-plugin.sh
|
||||
|
|
|
|||
|
|
@ -1,6 +1,6 @@
|
|||
# Azure AI Architecture Playground
|
||||
|
||||
**Target:** Lagre denne planen i `/Users/ktg/.claude/plugins/marketplaces/ktg-privat/plugins/ms-ai-architect/docs/playground-plan.md`
|
||||
**Target:** Lagre denne planen i `<plugin-root>/docs/playground-plan.md`
|
||||
|
||||
---
|
||||
|
||||
|
|
@ -10,7 +10,7 @@
|
|||
|
||||
**Mal:** Bygge et guidet arkitekturbeslutningsverktoy for Azure AI-plattformen, ved a bruke Playground-pluginen som generator og ms-ai-architect som kunnskapsbase. Prompt-outputen fra playground limes tilbake i Claude for a kjore `/architect`-kommandoer.
|
||||
|
||||
**Arbeidskatalog:** `/Users/ktg/.claude/plugins/marketplaces/ktg-privat/plugins/ms-ai-architect/` (ms-ai-architect plugin-mappen)
|
||||
**Arbeidskatalog:** `<plugin-root>/` (ms-ai-architect plugin-mappen)
|
||||
|
||||
---
|
||||
|
||||
|
|
|
|||
|
|
@ -2676,7 +2676,7 @@
|
|||
]
|
||||
},
|
||||
{
|
||||
"id": "agent-ecosystem-and-ktg-privat",
|
||||
"id": "agent-ecosystem-and-marketplace",
|
||||
"title": "Agent Ecosystem and Plugin Marketplace Patterns",
|
||||
"description": "Bygging av agentekosystemer, plugin-markeder, third-party-integrasjoner og distribusjon av agentplugins.",
|
||||
"subtopics": [
|
||||
|
|
|
|||
|
|
@ -1073,4 +1073,4 @@ Dette området utvikler seg raskt. Anbefalt re-verification:
|
|||
|
||||
---
|
||||
|
||||
*Denne kunnskapsreferansen er sist oppdatert 2026-02-04 av Cosmo Skyberg, Microsoft AI Solution Architect. For spørsmål eller korreksjoner, kontakt via Linear issue tagging `🏛️ ARCHITECT`.*
|
||||
*Denne kunnskapsreferansen er sist oppdatert 2026-02-04 av Cosmo Skyberg, Microsoft AI Solution Architect.*
|
||||
|
|
|
|||
|
|
@ -116,9 +116,13 @@ print(response.choices[0].message.content)
|
|||
- **Ingen per-token avgift**
|
||||
|
||||
**Eksempel (Ollama):**
|
||||
|
||||
> [!WARNING]
|
||||
> `curl | sh` laster ned og kjører remote kode direkte. For produksjonsbruk: last ned scriptet, inspiser det, kjør deretter: `curl -fsSL https://ollama.com/install.sh -o install.sh && sh install.sh`
|
||||
|
||||
```bash
|
||||
# Installér Ollama
|
||||
curl -fsSL https://ollama.com/install.sh | sh
|
||||
curl -fsSL https://ollama.com/install.sh | sh # gitleaks:allow
|
||||
|
||||
# Last ned Phi-4-mini
|
||||
ollama pull phi4
|
||||
|
|
@ -325,9 +329,13 @@ kubectl run -it --rm --restart=Never curl --image=curlimages/curl -- curl -X POS
|
|||
**Ollama** er et lightweight rammeverk for å kjøre LLMs og SLMs lokalt.
|
||||
|
||||
**Eksempel (on-premises):**
|
||||
|
||||
> [!WARNING]
|
||||
> `curl | sh` laster ned og kjører remote kode direkte. Inspiser scriptet før kjøring i produksjon.
|
||||
|
||||
```bash
|
||||
# Installér Ollama
|
||||
curl -fsSL https://ollama.com/install.sh | sh
|
||||
curl -fsSL https://ollama.com/install.sh | sh # gitleaks:allow
|
||||
|
||||
# Last ned Phi-4
|
||||
ollama pull phi4
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue