docs(nas): the copy off smaug exists, and the two runbooks that said otherwise say so

[Gerrrt]

What changed

Documentation only. Two runbooks said erebor/apps had no copy off smaug after the morning it got one.

• build-the-nas.md — the §7 status paragraph about the faulted disk said "§6.2 — still not done — is its step 2, because erebor/apps has no copy off this host"; §8's first bullet said "Its step 2 is §6.2" as if owed. Both now say the copy ran on 2026-09-20 (set 20260920T060234Z, jellyfin.db present, hashed on oracle, timer installed; erebor/apps is declared backed up in two files and backed up by nothing #484 shut on it) and that the tray can be touched. §6.2 itself is untouched.
• replace-the-nas-disk.md — the title line, the status block ("steps 2–6 are not done", "exists only on it"), the "zero copies off smaug" paragraph, the port-22 sentence, and the What is still open list all carried the same claim. Each is past tense with the date, step 2 gets a Done 2026-09-20, path A note, and the "§6.2 until path A has run" bullet leaves the open list.

Why

#563 (the §6.2 Done block) and #565 (the fault readings, with the "still not done" sentences) merged from different branches on the same morning, so the file disagreed with itself by lunchtime. Found during the 2026-09-20 issue triage. Refs #558.

Blast radius

None. No host, rule or script changes.

• No change to network segmentation or firewall rules
• No new port published to a VLAN that could not already reach the service
• No credential added outside secrets/*.sops.yaml

Verification

• make lint passes; scripts/check_docs.py passes (docs OK — 89 Prometheus + 18 Loki rules, 10 assertions)
• Deployed to the lab and confirmed working — not applicable, documentation only
• Docs updated

🤖 Generated with Claude Code