The datatable column is type number, so the human-formatted string
('20.96 KB') was rejected with a validation error (seen on the first
vaultwarden-org-export report). Format for display in the reporter,
not in the payload. Deployed copy in ~/.local/bin updated to match.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0133oyd1U9XxpLQU3UHMzUyy
Home Assistant's Synology DSM integration connects on 5001 with SSL
(verification off, self-signed cert).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Confirms bad RAM (memtest86+ single-bit flips) as root cause of the
2026-06-24 Btrfs corruption incident, adds fedora-upgrade.md and other
pending docs, and gitignores the doc-hygiene state directory.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Chrome crashed repeatedly with SIGBUS on launch. Root cause was a
corrupt data block in ~/.config/google-chrome/Profile 2/Favicons on a
single-copy Btrfs volume (csum failures correlated to the second with
each crash). A scrub found two further uncorrectable blocks (a Chrome
cache file and a gdm help file); both fixed. SSD SMART is clean and the
corruption is static across scrubs, pointing at a past RAM/bus event
rather than a failing drive.
Adds docs/incidents.md (dated incident log) and a CLAUDE.md pointer.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The old Hyperthrive DigitalOcean VPS (161.35.12.75) was deleted
2026-06-24 after all services migrated to the OVH VPS (verified live,
no DNS pointing at the old IP). Mark the section DELETED, retain
recovery notes as historical, and update the CLAUDE.md docs pointer.
Full migration record lives in ~/projects/server-ovh-prod-01-main/.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
SFTP uses a virtual root where Synology shared folders appear directly
under / — /volume1/... does not exist in SFTP space. The repo lives at
/homes/jared/Backups/fedora-workstation, not /home/Backups/... .
Also removes the pre-flight reachability check (HTTP curl, previously SSH)
that has been the source of every Synology skip since May 24. Restic
handles unreachable SFTP endpoints gracefully; no gate needed.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Check that /volume1/backup is actually reachable and readable, rather
than just testing SSH connectivity. This prevents false positives where
SSH works but the backup destination is offline or unmounted.
Added comprehensive operational documentation for Synology DS218plus
access, backup paths, and restic commands. Added VPS inventory with
notes on Hyperthrive DigitalOcean recovery and OVH migration plan.
Updated CLAUDE.md to reference new docs, expand known workarounds
with cursor and NVIDIA DRM details, and clarify hardware status.
12-task plan covering Ruby notifier script (TDD, Sandi Metz OOP),
n8n datatable setup, both workflows, and backup.sh integrations
for desktop and OVH VPS.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Two-workflow n8n system: webhook receiver records backup run results
into a datatable; weekly Friday reporter compares against expected
sources and sends green/yellow/red health digest via MailPace.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Add keyboard documentation including README and JSON keymap exports
from the K15 Pro. Useful for reference when reconfiguring via
launcher.keychron.com or troubleshooting firmware issues.
Refresh cheatsheet.md with tested DMS keybindings and actual working
setup (e.g. SUPER+T for terminal, DMS spotlight, lxpolkit instead of
broken hyprpolkitagent). Update plan.md to replace hyprpolkitagent
with lxpolkit (Qt 6 ABI issue). Add detailed phase 1-3 progress notes
including lxpolkit workaround, hyprland.conf + hyprpaper verification,
and dotfiles stow integration.
jared