<!-- sid:test-version-1773862032 ws:magitek-ops pn:1 layer:L2-FTS type:initial ms:707 -->
[Syncrovanis v0.2.1] onboarding | L2-FTS via KB-FTS5 | magitek-ops | domener: claude/agents/expert-training-syncrovanis claude/agents/agent-meta-suitecrm claude/agents/expert-training-ops | 707ms
Layers: L0:miss|L1:miss|L2:hit|L3:skip(not-enabled)

**[Guardrail] expert-training-syncrovanis: Expert File Structure** _/home/heine/.claude/agents/expert-training-syncrovanis.md_
```
experts/
  CURRENT.md                              → symlink to latest version
  EXPERT-syncrovanis-v*.md                — Main engine expert
  EXPERT-syncrovanis-competitive-*.md     — Competitive analysis
```

**Symlink convention:** `CURRENT.md` ALWAYS points to the latest version. NEVER create raw `CURRENT.md` files.
**[Workflow] expert-training-syncrovanis: Expert Training Agent — Syncrovanis** _/home/heine/.claude/agents/expert-training-syncrovanis.md_
**Purpose:** Create and maintain expert knowledge files for the Syncrovanis context delivery engine, its monitoring dashboard, and the classification pipeline.

**Workspace:** `/var/www/syncrovanis/`
**Expert root:** `experts/`

---
**[Workflow] expert-training-syncrovanis: What Syncrovanis Is** _/home/heine/.claude/agents/expert-training-syncrovanis.md_
Dynamic context delivery engine for Claude Code. Two codebases:

**Engine (Bash):**
- Hook pipeline: `engine/hooks/user-prompt-context-engine.sh`
- Classification: `engine/lib/classify.sh` (L1 keyword, L2 FTS, L3 LLM)
- KB query: `engine/lib/kb-query.sh` (`kb_query_rich()` returns JSON)
- State: `engine/lib/state.sh` (session state, delta detection)
- Workspace detection: `engine/lib/workspace-detect.sh`
- Quality test: `engine/test/quality-test.sh`

**Dashboard (Python/Flask):**
- Flask app:...
**[Workflow] expert-training-syncrovanis: Knowledge Domains to Cover** _/home/heine/.claude/agents/expert-training-syncrovanis.md_
Expert files for Syncrovanis should document:

1. **Pipeline architecture** — Hook flow, classification layers, KB query, context delivery
2. **Classification system** — L1 keywords, L2 FTS, L3 LLM, provider constraints
3. **KB integration** — `kb_query_rich()`, JSON format, artisan footer workaround, dedup
4. **State management** — Session state, delta detection, onboarding vs continuation
5. **Cross-workspace behavior** — How configs differ, testing strategies
6. **Dashboard architecture** —...
**[Workflow] expert-training-syncrovanis: Overview** _/home/heine/.claude/agents/expert-training-syncrovanis.md_
[What this covers and why it matters]
**[Guardrail] agent-meta-suitecrm: Self-Awareness (CRITICAL)** _/home/heine/.claude/agents/agent-meta-suitecrm.md_
When user asks to "fix", "update", or "improve" `/agent-meta` in this workspace:

**This agent's own config lives in these files — edit them directly:**
1. `~/.claude/agents/agent-meta-suitecrm.md` — THIS FILE (specialized variant, primary procedures)
2. `~/.claude/agents/agent-meta/context-suitecrm-dev.md` — Workspace data (agents, commands, KB, experts)
3. `~/.claude/agents/agent-meta.md` — Router (only touch if routing logic needs updating)
4. `~/.claude/agents/agent-meta-base.md` — Generic...
**[Guardrail] agent-meta-suitecrm: 1. Agent: `suitecrm-specialist` (primary domain agent)** _/home/heine/.claude/agents/agent-meta-suitecrm.md_
**File:** `~/.claude/agents/suitecrm-specialist.md` (~190 lines, slim router + behavior)

**Knowledge files** (8 stk, in `~/.claude/agents/suitecrm-specialist/`):

| File | Domain | Update when... |
|------|--------|----------------|
| `knowledge-architecture.md` | Dual-layer system, file structure, DB | Architecture changes, new layers discovered |
| `knowledge-extensions.md` | Extension framework, QR&R, vardefs, layoutdefs | New extension patterns, build process changes |
|...
**[Guardrail] agent-meta-suitecrm: MCP Feedback Loop** _/home/heine/.claude/agents/agent-meta-suitecrm.md_
**Quick Fetch — load discovery tool schemas before investigating MCP gaps:**
```
# SuiteCRM MCP — test actual tool availability
ToolSearch("select:mcp__suitecrm__get_module_fields,mcp__suitecrm__list_records,mcp__suitecrm__get_relationships,mcp__suitecrm__get_view_layout,mcp__suitecrm__search_labels,mcp__suitecrm__deploy_metadata_file,mcp__suitecrm__cache_clear")

# Source code navigation (remote Serena on SuiteCRM...
**[Tool usage] agent-meta-suitecrm: 5. Explore Reports** _/home/heine/.claude/agents/agent-meta-suitecrm.md_
```
coordination/explore/
├── EX-0013 suitecrm-mcp-server
├── EX-0016 suitecrm-agent-strategy
├── EX-0018 suitecrm-complete-inventory (NOTE: duplicate sequence number)
├── EX-0018 suitecrm-extension-framework
├── EX-0019 suitecrm-ui-customization-docs
└── EX-0020 suitecrm-ui-customization
```
**[Guardrail] expert-training-ops: CRITICAL: Location-Based Directory Structure** _/home/heine/.claude/agents/expert-training-ops.md_
Infrastructure experts are organized by **physical location**, NOT flat domains:

```
coordination/experts/operations/magitek-server-ops/
├── CURRENT.md                  # Master index (lists ALL sub-experts with location)
├── kontoret/                   # 172.20.0.x network
│   ├── proxmox/                # Proxmox hosts (px1)
│   ├── network/                # pfSense, switches, APs
│   └── services/               # NPM, legacy-tls-proxy, ssh-server, reportmaker
├── hjemme/                    ...
**[Guardrail] expert-training-ops: Step 3: Check Backup Status** _/home/heine/.claude/agents/expert-training-ops.md_
**CRITICAL — from POSTMORTEM: backup is configured at Proxmox DATACENTER level, NOT inside VM/CT!**

```
ssh {proxmox_host} "cat /etc/pve/jobs.cfg"
```

Check if this VM/CT's VMID appears in any backup job's `vmid` list.
**[Guardrail] expert-training-ops: FORBIDDEN Content:** _/home/heine/.claude/agents/expert-training-ops.md_
- Code blocks >3 lines
- External URLs
- "Example:" or "Here's how..." sections
- Meta-information (confidence, tokens)
- Human-friendly fluff
**[Guardrail] expert-training-ops: MCP Tools (Always Available)** _/home/heine/.claude/agents/expert-training-ops.md_
All MCP servers are active (deferred loading — 0 cost until first use). For expert file creation:

**Quick Fetch — load these for infrastructure work:**
```
# Code/config navigation
ToolSearch("select:mcp__serena__find_symbol,mcp__serena__get_symbols_overview,mcp__serena__find_referencing_symbols,mcp__serena__search_for_pattern")

# Database inspection
ToolSearch("select:mcp__mysql__execute_sql")

# Documentation...
**[Guardrail] expert-training-ops: Symlink Rules** _/home/heine/.claude/agents/expert-training-ops.md_
- CURRENT*.md MUST be symlinks, NEVER raw files
- One file = One symlink (no redundancy)
- Verify with `file CURRENT-{name}.md` → must say "symbolic link"

MCP: proxmox-hjemme, proxmox-pmox2, proxmox-pmox10, proxmox-pmox15, proxmox-px1, proxmox-px3, docker, serena, context7, dam-sqlite

Onboardet av Syncrovanis | L2-FTS via KB-FTS5 | 707ms

INSTRUKS: Start ditt neste svar med NØYAKTIG disse to linjene (kopier ordrett, inkludert markdown-link):
Onboardet av Syncrovanis | L2-FTS via KB-FTS5 | 707ms
[Syncrovanis v0.2.1](.claude/syncrovanis-context.md) initial | L2-FTS via KB-FTS5 | magitek-ops | claude/agents/expert-training-syncrovanis claude/agents/agent-meta-suitecrm claude/agents/expert-training-ops | 707ms