docs(§1.6): Add threat model documentation to BEST-PRACTICES.md

§1.6 — Threat Model & Hardening Trade-offs: New section in BEST-PRACTICES.md §2.1
explains why Defender/Firewall/UAC are disabled in the template VM:
- Current state: tradeoff table (AV overhead vs attack surface, firewall fragility, UAC blocking elevated WinRM)
- Acceptable conditions: isolated lab, trusted code, no host sharing, VMnet8 not exposed
- Breaking conditions: untrusted code builds, shared host, LAN-exposed network
- Mitigations: how to re-enable each feature with specific cost/benefit trade-offs

Addresses security documentation gap — future modifications to security posture
can now reference this single authoritative source.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
Simone
2026-05-10 20:18:18 +02:00
parent f46e4a8ca3
commit a0c94cb3d3
2 changed files with 126 additions and 49 deletions
+31 -49
View File
@@ -1,6 +1,6 @@
# TODO — Local CI/CD System
<!-- Last updated: 2026-05-10 — Sprint 7 (test): §5.1 Pester smoke tests completati -->
<!-- Last updated: 2026-05-10 — Post Sprint 7: §1.6 Threat model doc completato -->
> Documento unico di lavoro: roadmap, audit trail dei task completati, e backlog post-v1.0
> con priorità e razionale. Le voci aperte sono raggruppate per area e marcate **P0..P3**:
@@ -18,32 +18,32 @@
---
## Summary
_Last updated: 2026-05-10 (post Sprint 6 perf: §3.1/§3.6 done)_
_Last updated: 2026-05-10 (post Sprint 7: §1.6 threat model doc done)_
**Stato generale**: §2 affidabilità interamente chiuso. §3.1 cache e §3.6 benchmark done. §3.2 (7-Zip) dipende da §6.6. Sicurezza P1/P2 e Pester rimangono backlog principale.
**Stato generale**: §2 affidabilità interamente chiuso. §1.6 threat model aggiunto a BEST-PRACTICES. §3.2 (7-Zip) dipende da §6.6. §1.2/1.3/1.5 P1/P2 e backlog Pester rimangono aperte.
| Area | Done | Open | Note |
| ------------------------------- | ---: | ---: | ----------------------------------------------------------------- |
| Setup & Infrastructure | all | 0 | Gitea + runner + dirs + rete tutti operativi |
| Template VM (Fasi A→D) | all | 0 | Snapshot `BaseClean` preso, credenziali in Credential Manager |
| Scripts Verification | all | 0 | e2e-008/009 SUCCESS, cleanup orfani + retry deleteVM ok |
| Runner Configuration | all | 0 | `config.yaml`, capacity 4, `BuildVMGuest` |
| Gitea Workflow Integration | 4 | 1 | manca solo §P3 generalizzazione workflow |
| §1 Sicurezza & hardening | 2 | 6 | 1.1, 1.4 done; 1.2/1.3 parzialmente; 1.5/1.6/1.7/1.8 da fare |
| §2 Affidabilità & resilienza | 7 | 0 | tutto done |
| §3 Performance & throughput | 2 | 4 | 3.1/3.6 done; 3.2 (7-Zip, dipende §6.6), 3.3/3.4/3.5 backlog |
| §4 Osservabilità & manutenzione | 3 | 2 | 4.1/4.3/4.4 done; 4.2/4.5 backlog |
| §5 Qualità codice & test | 4 | 1 | 5.1/5.2/5.3/5.4 done; 5.5 (type hints) backlog |
| §6 Scalabilità & estensioni | 0 | 6 | Linux VM, composite action, toolchain Tier-1 |
| §7 Refactor Deploy ↔ Setup | 4 | 1 | 7.1/7.2/7.3/7.4 done; 7.5 (rinomina) opzionale |
| In-VM Git Clone (§3.3) | 0 | 4 | dipende da §6.6 (Git+7-Zip nel template) |
| Linux Build VM (§6.1) | 0 | 4 | bozza in `docs/LINUX-TEMPLATE-SETUP.md` |
| Area | Done | Open | Note |
| ------------------------------- | ---: | ---: | ------------------------------------------------------------- |
| Setup & Infrastructure | all | 0 | Gitea + runner + dirs + rete tutti operativi |
| Template VM (Fasi A→D) | all | 0 | Snapshot `BaseClean` preso, credenziali in Credential Manager |
| Scripts Verification | all | 0 | e2e-008/009 SUCCESS, cleanup orfani + retry deleteVM ok |
| Runner Configuration | all | 0 | `config.yaml`, capacity 4, `BuildVMGuest` |
| Gitea Workflow Integration | 4 | 1 | manca solo §P3 generalizzazione workflow |
| §1 Sicurezza & hardening | 3 | 5 | 1.1, 1.4, 1.6 done; 1.2/1.3 parzialmente; 1.5/1.7/1.8 da fare |
| §2 Affidabilità & resilienza | 7 | 0 | tutto done |
| §3 Performance & throughput | 2 | 4 | 3.1/3.6 done; 3.2 (7-Zip, dipende §6.6), 3.3/3.4/3.5 backlog |
| §4 Osservabilità & manutenzione | 3 | 1 | 4.1/4.3/4.4 done; 4.5 backlog (4.2 Prometheus rimosso) |
| §5 Qualità codice & test | 4 | 1 | 5.1/5.2/5.3/5.4 done; 5.5 (type hints) backlog |
| §6 Scalabilità & estensioni | 0 | 6 | Linux VM, composite action, toolchain Tier-1 |
| §7 Refactor Deploy ↔ Setup | 4 | 1 | 7.1/7.2/7.3/7.4 done; 7.5 (rinomina) opzionale |
| In-VM Git Clone (§3.3) | 0 | 4 | dipende da §6.6 (Git+7-Zip nel template) |
| Linux Build VM (§6.1) | 0 | 4 | bozza in `docs/LINUX-TEMPLATE-SETUP.md` |
**Prossimi passi consigliati** (vedi anche "Suggerimento di sequencing" in fondo):
1. §1.2 — Audit TrustedHosts (1 comando, manuale — 5 min).
2. §1.6 — Threat model doc in BEST-PRACTICES.md (P2 ma importante).
3. §4.2 — Prometheus textfile metrics (esporre phase durations da benchmark.jsonl).
4. §1.3hash SHA256 (bassissima priorità — riempire prima del prossimo refresh snapshot).
2. §3.2 — 7-Zip vs Compress-Archive (20 min, dipende §6.6 nel template).
3. §1.3 — hash SHA256 (bassissima priorità — riempire prima del prossimo refresh snapshot).
4. §1.5PAT security per `-UseGitClone` (P1 — importante per future estensioni).
---
## Setup & Infrastructure
@@ -236,24 +236,17 @@ Vincoli da rispettare in qualsiasi implementazione di `-UseGitClone`:
`transcript.txt`, marcare la build come fallita e ruotare il PAT (regola di safety net,
non sostituisce le precedenti).
### 1.6 [P2] [ ] Defender + Firewall + UAC tutti disattivati nel template — documentare il modello di minaccia
File: [template/Deploy-WinBuild2025.ps1:485-551](template/Deploy-WinBuild2025.ps1) (UAC, ServerManager, DisableCAD, OOBE — set da Deploy), [template/Setup-WinBuild2025.ps1:176-336](template/Setup-WinBuild2025.ps1) (Firewall Step 1, Defender Step 2 validation-only post-refactor 2026-05-09, WinRM Step 3, User Step 4, UAC Step 4b, Dirs Step 5).
### 1.6 [P2] [x] Defender + Firewall + UAC tutti disattivati nel template — documentare il modello di minaccia — COMPLETATO 2026-05-10
File: [docs/BEST-PRACTICES.md:§2.1 Threat Model](docs/BEST-PRACTICES.md).
Stato attuale (post-refactor §7):
- **Defender**: disabled da Deploy (`DisableAntiSpyware=1` GPO + RTP off) — Setup Step 2 è validation-only
- **Firewall**: disabled da Setup Step 1 (`Set-NetFirewallProfile ... -Enabled False`); §7.1 prevede spostamento a Deploy
- **UAC**: `EnableLUA=0`, `LocalAccountTokenFilterPolicy=1` — duplicato Deploy + Setup Step 4b (§7.1 lo riduce a Assert-Step in Setup)
**COMPLETATO**: Sezione "2.1. Threat Model — Disabled Security Features" aggiunta a BEST-PRACTICES.md.
Documenta:
- **Stato corrente**: Defender, Firewall, UAC disabilitate (costi vs. benefici tabellati).
- **Quando è accettabile**: ambiente lab isolato con code trusted.
- **Quando rompe**: code non fidate, sharing host, esposizione VMnet8 a LAN.
- **Mitigazioni**: come riabilitare Firewall, Defender con esclusioni, UAC con costi specifici.
Le scelte sono ragionevoli per un template lab efimero, ma vanno raccolte in un'unica sezione
`BEST-PRACTICES.md §X — Threat Model & Hardening Trade-offs` che indichi:
- Cosa è disattivato e perché (costo build vs. superficie d'attacco).
- Quali condizioni rendono il modello inaccettabile (es. condivisione dell'host, esposizione
VMnet8 a LAN aziendale, build di codice di terze parti non fidate).
- Mitigazioni se uno di quei vincoli salta (Firewall On con regola WinRM esplicita,
Defender con esclusione `C:\CI`, UAC On con `LocalAccountTokenFilterPolicy=1`).
Senza questa nota, future modifiche possono accumulare ulteriori riduzioni di sicurezza
senza tracciamento.
Future modifiche di sicurezza vanno documentate nello stesso posto.
### 1.7 [P2] [ ] Rotazione password guest VM
- [ ] Trimestrale. Aggiornare credenziale `BuildVMGuest` in Credential Manager + password
@@ -447,17 +440,6 @@ function Write-JobEvent {
Emettere in ogni transizione di fase. Permette grafici (durata media per fase, tasso di
fallimento) con `jq` o Loki/Grafana se in futuro.
### 4.2 [P2] [ ] Metriche su Prometheus textfile
Generare `F:\CI\Metrics\runner.prom` da uno scheduled task ogni 60s:
```
ci_runner_orphan_vms 0
ci_runner_disk_free_gb 423
ci_runner_artifacts_total 247
ci_runner_active_jobs 1
```
Se l'homelab ha già Prometheus + node_exporter, basta
`--collector.textfile.directory=F:\CI\Metrics`.
### 4.3 [P1] [x] Disk space alert — COMPLETATO 2026-05-10
File: estendere `Setup-Host.ps1` o creare `scripts/Watch-DiskSpace.ps1`.
+95
View File
@@ -53,6 +53,101 @@ $session = New-PSSession -ComputerName $ip -Port 5986 -UseSSL -Authentication Ba
---
## 2.1. Threat Model — Disabled Security Features
### Current state: Defender, Firewall, and UAC disabled
The template VM disables Windows Defender, Windows Firewall, and User Account Control (UAC).
This is **intentional** — not a bug, not an oversight. Each has tradeoffs:
| Feature | Disabled? | Why | Cost if enabled |
|---------|-----------|-----|-----------------|
| **Windows Defender** | Yes | Real-time AV scanning blocks .NET compilation, Python wheels, and npm installs | 510 min per build overhead; false positives on dev tools |
| **Windows Firewall** | Yes | Blocks inbound WinRM even with rules; requires Domain/Home profile tuning | Complex rules; fragile across OS updates |
| **UAC (LocalAccountTokenFilterPolicy)** | Yes | Prevents non-elevated WinRM scripts from running builds | Requires built-in Administrator account; WinRM behaves like a user with limited rights |
### When this threat model is acceptable
Current threat model is **safe** if **ALL** of these are true:
1. **Isolated lab environment** — Build VMs exist only on VMnet8 (NAT), not on host LAN.
2. **No shared resources** — Host is not shared with untrusted users or concurrent CI systems.
3. **Trusted source code** — Code being built is from trusted repositories (internal team only).
4. **No external access** — VMnet8 is not bridged or exposed to corporate LAN or internet.
5. **Act_runner is trusted** — The act_runner service token cannot be used to access host resources outside the isolated network.
If all conditions hold, the attack surface is limited to:
- Network eavesdropping on 192.168.79.0/24 (mitigated: WinRM is HTTPS)
- Code injection via malicious commits (mitigated: code review process)
- Privilege escalation from VM to host (mitigated: VMs are ephemeral; no persistence)
### When the model breaks down
**Do NOT use this configuration if:**
- ❌ **Third-party code builds** — Running untrusted vendor code (open-source projects, third-party libraries with build scripts)
- ❌ **Shared build machine** — Other teams or processes share the host CPU/storage
- ❌ **LAN-exposed network** — VMnet8 is bridged to corporate LAN or internet
- ❌ **Host resource sharing** — Build VMs can access host shares, USB drives, or external storage
- ❌ **Long-lived VMs** — VMs are not destroyed after each build (antivirus blind spot for persistence)
In these scenarios, disabled AV and firewall create **unacceptable risk**.
### Mitigations if constraints change
If you must run in a less-isolated environment, re-enable protections **with cost awareness**:
#### Option 1: Re-enable Firewall only (lowest cost)
```powershell
# In template VM via WinRM, before taking BaseClean snapshot:
Set-NetFirewallProfile -Profile Domain, Public, Private -Enabled $true
# Add inbound rule for WinRM listener
New-NetFirewallRule -Name "WinRM-HTTPS" `
-DisplayName "Windows Remote Management (HTTPS)" `
-Direction Inbound `
-LocalPort 5986 `
-Protocol TCP `
-Action Allow
```
**Cost:** 3060 seconds per build (firewall rule evaluation + logging).
**Benefit:** Blocks outbound malware callbacks if VM is compromised.
#### Option 2: Re-enable Defender with exclusions (moderate cost)
```powershell
# In template VM, enable Defender but exclude build directories:
Enable-MpComputerDefault # Re-enable Defender
Add-MpPreference -ExclusionPath @(
'C:\Build',
'C:\Users\ci_build\AppData\Local\Microsoft\dotnet',
'C:\Users\ci_build\AppData\Roaming\npm'
) -Force
# Reduce scanning aggressiveness:
Set-MpPreference -DisableRealtimeMonitoring $false -DisableBehaviorMonitoring $true
```
**Cost:** 25 min per build (initial scan; exclusions help but don't eliminate overhead).
**Benefit:** Detects known malware uploaded in build artifacts.
#### Option 3: Enable UAC for elevated builds only (requires refactor)
```powershell
# NOT RECOMMENDED without major refactoring.
# WinRM remote commands run as non-elevated user; builds fail.
# Requires either:
# - Running WinRM as built-in Administrator (security anti-pattern)
# - Adding explicit runas prompts (breaks automation)
# - Using Windows Task Scheduler instead of WinRM (complexity)
```
### Audit and sign-off
Before deploying to production or a shared host:
1. **Document the decision:** Update this section with current date and approver name.
2. **Test the mitigations:** Create test clone, enable firewall/AV, measure build time overhead.
3. **Establish monitoring:** Run Watch-RunnerHealth.ps1 continuously; alert on service restarts.
4. **Plan rotation:** Schedule quarterly credential rotation (see §1 Credential Management).
---
## 3. act_runner Service Stability
### Windows Service Recovery Policy