Fixes contradictions found during the CT 104 disk-full incident:
- Arcane: was DECOMMISSIONED 2026-05-26 + local, not "migrated to UNAS"
(volumes.md and storage.md disagreed).
- arr-stack: fully on UNAS incl. SQLite configs on services/arr-stack/;
media/downloads on media/ share. Radarr/Sonarr not deployed. Removed
stale "plan to migrate" rows pointing at local /opt/stacks/arr-stack.
- n8n: local (DB on CT 113), not UNAS.
- Garage: metadata is SQLite (not LMDB); keep off NFS; refreshed bucket
table with real sizes (82 GB, mostly WAL-G PG backups).
- Marked sunsets: karakeep, memos, litellm, daytona.
- crawl4ai: correct name mcp-crawl4ai, no mounts, cache-growth TODO.
- Added ZFS snapshot note: refquota hides snapshot bloat from container df;
pre-tier1-backups snapshot held 97.9 GB.
- Add RESUME.md for cross-session continuity (open items, key state, constraints)
- claude-max-bridge: implement /v1/responses (OpenAI Responses API) with
previous_response_id chaining via server-side history injection into system prompt.
Root cause of "session already in use": CLI leaves JSONL in un-resumable
"dequeued" state after each --print run; fix avoids session reuse entirely.
Also fixed: assistant content must be array-of-blocks not plain string (silent
JS crash otherwise).
- LiteLLM: add pass_through_endpoints for /v1/responses → claude-max-bridge
- Storage, volumes, architecture docs reconciled (Vaultwarden → local zfs,
Pocket-ID backup, WAL-G fix, apps/ decommission, Nextcloud CIFS→NFS)
- Add ideas/, proxmox-memory-audit.md, llm-benchmark.md (new docs this session)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
fkrebs