Initial infrastructure mono-repo scaffold

Consolidates nixos, docker, raspi, and debian-configuration into a single
infrastructure-as-code repo. Includes:

- ansible/: full inventory + proxmox-hardening, freeipa, and raspberrypi
  roles (converted from debian-configuration bash scripts)
- terraform/: Proxmox VMs, Dynu DNS, Pi-hole (decommissioned stub),
  Docker container catalog — migrated from docker/infrastructure/terraform/
- stacks/docker/, stacks/raspi/, nixos/: placeholder READMEs pending
  git subtree population (see implementation plan)
- docs/: internal MkDocs site with architecture, network topology, runbooks,
  and drift-detection guide; external sanitized site
- scripts/: drift-detect.sh, docs-build.sh, install-hooks.sh, check-secrets.sh
- CI: secret-scan (push/PR), drift-detect (daily), docs-build (on change)
- Pi-hole removed throughout — DNS is FreeIPA, DHCP is router

See docs/internal/implementation-plan.md for the phased rollout after
pushing to Gitea.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UvNjoxTWEDkhXsd1Dq2ETP

docs/internal/runbooks/drift-detection.md

@@ -0,0 +1,64 @@
+# Runbook: Drift Detection
+
+How to interpret and respond to drift detected by the CI pipeline or local script.
+
+## What is drift?
+
+**Terraform drift** — a Proxmox VM, DNS record, or Pi-hole entry exists in a different
+state than what's declared in the Terraform workspace. Common causes:
+- A VM was resized or reconfigured manually in the PVE web UI
+- A DNS record was added directly via Dynu or Pi-hole admin
+- A VM was deleted outside of Terraform
+
+**Ansible drift** — a host's configuration differs from what the Ansible role declares.
+Common causes:
+- A config file was edited by hand on the host
+- A package was installed/removed manually
+- A service was stopped/disabled without updating the role
+
+## CI workflow
+
+The `drift-detect` workflow runs daily at 06:00 AEST. If drift is detected, a Gotify
+notification is sent to the push notification dashboard. The step summary in Gitea CI
+contains the full `terraform plan` diff.
+
+## Responding to Terraform drift
+
+```bash
+# See exactly what changed
+cd terraform/proxmox
+terraform plan
+
+# Option A: The manual change was intentional — update the .tf file to match reality,
+# then run terraform apply to reconcile state
+vim proxmox/docker.tf  # update the resource to match current state
+terraform apply
+
+# Option B: The manual change was accidental — apply to restore declared state
+terraform apply  # will revert the manual change
+```
+
+## Responding to Ansible drift
+
+```bash
+# See exactly what would change
+cd ansible
+ansible-playbook playbooks/site.yml --check --diff --limit <drifted_host>
+
+# Apply to restore declared state
+ansible-playbook playbooks/site.yml --limit <drifted_host>
+```
+
+## Required Gitea secrets
+
+The following secrets must be configured in the Gitea repo for CI drift detection to work:
+
+| Secret | Used by |
+|--------|---------|
+| `PROXMOX_ENDPOINT` | Terraform proxmox workspace |
+| `PROXMOX_API_TOKEN_ID` | Terraform proxmox workspace |
+| `PROXMOX_API_TOKEN_SECRET` | Terraform proxmox workspace |
+| `DYNU_API_KEY` | Terraform dns workspace |
+| `ANSIBLE_SSH_KEY` | Ansible drift check |
+| `GOTIFY_URL` | Drift notifications |
+| `GOTIFY_TOKEN` | Drift notifications |

docs/internal/runbooks/new-host.md

@@ -0,0 +1,59 @@
+# Runbook: Provisioning a New Host
+
+Steps for adding a new machine to the infrastructure.
+
+## NixOS host on Proxmox
+
+```bash
+# 1. Use the PXE boot menu or installer ISO
+#    The auto-installer at pxe-boot.sweet.home presents all flake targets.
+
+# 2. Add to nixos/ flake if a new build-type is needed
+#    Otherwise pick the closest existing target (e.g. proxmox-minimal)
+
+# 3. Add SSH host key to clan vars (enables SOPS decryption)
+cd nixos
+./scripts/sync-host-keys.sh <target>
+
+# 4. Create the VM/LXC via Terraform
+cd ../terraform/proxmox
+# Add a resource block in the appropriate .tf file
+terraform plan
+terraform apply
+
+# 5. Deploy NixOS
+nixos-rebuild switch --flake ./nixos#<target> --target-host <hostname>
+```
+
+## Non-NixOS host (Proxmox VM)
+
+```bash
+# 1. Create VM in Proxmox (manually or via terraform)
+# 2. Add to ansible/inventory/hosts.yml with correct group
+# 3. Add SSH key, verify connectivity:
+ansible-playbook ansible/playbooks/ping.yml --limit <hostname>
+# 4. Apply the role:
+ansible-playbook ansible/playbooks/<role>.yml --limit <hostname> --check --diff
+ansible-playbook ansible/playbooks/<role>.yml --limit <hostname>
+```
+
+## Adding Ansible inventory entry
+
+```yaml
+# ansible/inventory/hosts.yml — under the appropriate group:
+new-host.sweet.home:
+  ansible_user: wayne
+  # any role-specific variables
+```
+
+## IPA enrollment (for any new host)
+
+```bash
+# On the new host (after OS install):
+sudo ipa-client-install \
+  --server=domain-controller.sweet.home \
+  --domain=sweet.home \
+  --realm=SWEET.HOME \
+  --principal=admin \
+  --mkhomedir
+```

docs/internal/runbooks/proxmox-hardening.md

@@ -0,0 +1,52 @@
+# Runbook: Proxmox VE Hardening
+
+Applies Stage 1 hardening to a fresh Proxmox VE node using the `proxmox-hardening` Ansible role.
+
+## Pre-conditions
+
+- Fresh PVE install with root SSH access
+- SSH key already added to root's `authorized_keys`
+- Host added to `ansible/inventory/hosts.yml` under the `proxmox` group
+
+## Steps
+
+```bash
+# 1. Verify connectivity
+cd ansible
+ansible-playbook playbooks/ping.yml --limit <hostname>
+
+# 2. Dry-run to preview changes
+ansible-playbook playbooks/proxmox.yml --limit <hostname> --check --diff
+
+# 3. Review the diff output, then apply
+ansible-playbook playbooks/proxmox.yml --limit <hostname>
+
+# 4. Verify hardening state
+ansible-playbook playbooks/proxmox.yml --limit <hostname> --tags audit --check
+```
+
+## What gets applied
+
+1. **APT repos** — enterprise source disabled, no-subscription source enabled
+2. **SSH hardening** — `PermitRootLogin prohibit-password`, `PasswordAuthentication no`, fail2ban
+3. **Datacenter firewall** — default-deny inbound; SSH (22) and web UI (8006) from management CIDR only
+4. **Unattended upgrades** — security-only origins, no auto-reboot
+5. **Named admin user** — `<proxmox_admin_username>@pve` with Administrator role (one-time)
+6. **IPA sudo** — `admins` group gets NOPASSWD sudo for shell and PVE tools
+7. **Subscription nag** — cosmetic patch to remove the nag dialog
+
+## Variables to set per-host
+
+In `ansible/inventory/hosts.yml`:
+```yaml
+pve1.sweet.home:
+  proxmox_mgmt_cidr: "192.168.2.0/24"
+  proxmox_admin_username: wayne
+```
+
+## After hardening
+
+- Log into the PVE web UI as the named admin, change the initial password
+- Enable TOTP/2FA for both the named admin and root@pam
+- Verify with the audit script: `cd /home/wayne/repos/debian-configuration && proxmox/scripts/audit.sh`
+  (will migrate to an Ansible audit task in a future iteration)