/autotask exists nowhere as a Claude command (no .claude/commands/autotask.md in
the repo; only a Grok skill by that name). It was a GURU-5070-specific artifact in
the provisional manifest and produced a spurious RED on every other machine. Removing
it clears the false FAIL fleet-wide.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Adds .claude/scripts/migrate-to-submodules.sh — self-contained, distributable by
raw URL since old clones can't pull. Detects compliance (history merge-base vs
origin, RECLONE.md+submodule offline fallback); leaves compliant clones untouched;
otherwise re-clones AND recovers the gitignored per-machine state a clone never
carries (identity.json, settings.local.json, .mcp.json, grepai, per-project
.env/.venv/.attachments), surfaces stranded unpushed commits, and FLAGS large
purged data for manual move (never re-imports it into git). Closes RECLONE.md's
"recover any uncommitted work" gap that stranded identity.json + the discord-bot venv.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Codifies the scan-first/data-driven workflow proven on Cascades (where the baked-in
non-DFS bias picked the congested channels and a data-driven DFS plan halved 5GHz retry):
- NEW survey-report.py: rolls survey-collect JSON into the fleet per-channel/per-band-group
measured busy% table + cleanest/dirtiest ranking + a suggested clean 40MHz palette. The
decision-driver that was missing (we built it by hand).
- channel-plan.sh: na palette is now DATA-DRIVEN, not hardcoded non-DFS. Adds --channels
(explicit palette) + --dfs ok|avoid|only; default considers ALL 40MHz primaries and lets
measured busy% choose. Adds load-balancing + a local-search pass -> strong co-channel to 0.
- survey-collect.sh: per-AP "cleanest" report no longer pre-filters out DFS (DFS is usually
cleanest here); marks DFS with *, points at survey-report.
- SKILL.md: documents the mandatory scan -> survey-report -> channel-plan --channels -> apply
-> validate order + the Cascades lesson.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Did it right this time: completed the full channel survey (74/74) FIRST, let the
data choose. Survey proved DFS channels are 4-5x cleaner here (2-3% busy) than
non-DFS (149/157 = 12-28%, the property's worst). Per Howard: built the plan on
the 8 clean DFS 40MHz blocks (52/60/100/108/116/124/132/140), per-AP locally-
cleanest + neighbor graph-colored -> 0 co-channel, 3.5% avg busy. Applied to 72
non-mesh APs (width 40 too); mesh excluded; voice nudged back to 5GHz.
VALIDATED: 5GHz retry 8.7->3.8 avg (-56%), median 8.2->2.1 (-74%); 2.4 ~baseline;
satisfaction median 99; voice 31/31 (17 Poly on 5GHz, 3 coverage-cases on 2.4);
all 72 APs holding DFS, 0 radar vacates.
Kept tonight: 2b (2.4 power) + DFS plan + BSS-transition. 6GHz still WPA3-blocked.
auto_upgrade still OFF. Follow-up: recurring dfs-check radar monitor.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Correction to earlier "deferred" report: after Howard pushed (5GHz needs fixing
regardless of 6GHz), I attempted width40 + non-DFS channel plan autonomously.
It did NOT validate live: 5G retry flat (8.7->8.4), 2.4 retry up (12->16) from
voice phones scattering to 2.4. ROOT CAUSE: the non-DFS channels here (149/157)
carry the heaviest EXTERNAL interference while DFS was cleaner -> forcing non-DFS
traded clean DFS for congested non-DFS. Rolled 5GHz back to baseline (channel+80MHz).
Kicked the 8 stuck Poly phones -> 6 back to 5GHz (rest are coverage-gap rooms).
End state recovered: satisfaction 98.4/med99, voice 31/31. Kept: 2b (2.4 power)
+ BSS-transition. 5GHz unchanged from start. auto_upgrade left OFF.
Doing 5GHz right needs the per-channel survey (choose channels by real cleanliness,
not non-DFS policy), reconsider non-DFS-only, 6GHz unblock (WPA3), band-steer voice.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Read-only Phase 0 baseline extended to floors 5/6 (MemCare). Findings:
- MemCare = same 3 diseases as 1-4 but untreated (2.4 at full power, not
over-thinned; all 5GHz on DFS+80MHz; min-RSSI off everywhere; 6GHz dark;
Shelby .218 stuck on 2.4 at Nurse Station).
- 5GHz static non-DFS channel-plan dry-run: co-channel pairs 19 -> 0 (kills
auto on all non-mesh APs; relieves AP 103/505 as fall-out).
- 2.4 1/6/11 re-color NEGATIVE right now (22 -> 28); defer until 2b restores
a stable Medium-power radio set.
- 7-day hour-of-day traffic: ~600 clients 24/7 (only ~10% swing); trough
01:00-04:00 MST. Change window decided: 2 AM start.
No changes applied. Survey stalled 68/74 (re-run before any 5GHz channel apply).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
--full recompile. Folds the 4 repo-rewrite-recovered docs into the article
(master plan, voice QoS design, voice-quality diagnostic, RF/voice session log)
while preserving all existing depth. Key corrections/additions:
- Voice VLAN 30 cutover now COMPLETE: 8 AudioCodes (.224-.231) added; prior
compile had them 0/8 pending. AudioCodes needed a physical power-cycle
(externally powered/PoE off; controller bounce is a no-op).
- Poly fleet is 28 not 22 (6 stragglers still off VOICE).
- Voice quality is an RF problem on the Poly WiFi phones, not the VLAN move
(per-phone diagnostic; Lauren .202 50% retry -> locked to AP 103; AP 103
5GHz saturated; coverage gaps rooms 515/210/204).
- 6 GHz dark root-caused (CSCNet not broadcasting 6g).
- Measured WAN1 upload ~522 Mbps -> voice QoS is insurance, RF is the fix.
- New Patterns: Voice QoS design, Network Optimization Master Plan,
Decisions-2026-06-18 (non-DFS only; no dedicated voice SSID).
- Active Work / History / HIPAA reconciled to the complete cutover.
Live Syncro unchanged (55.75 hrs, 0 tickets, 29 assets). Synthesis was
deliberate surgical enrichment (diff-reviewed), not a blind regenerate.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Refresh-mode update. Live Syncro values unchanged (55.75 prepaid hrs, 0 open
tickets as of 2026-06-18). Added the 4 docs recovered in c2e5f4f to sources:
RF/voice optimization session log, network-optimization master plan, phase-1
voice QoS design, and the 2026-06-18 voice-quality diagnostic.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The 2026-06-18 repo restructure (history rewrite + project->submodule split)
dropped these 4 Cascades files from the new clone. Copied byte-identical from
the pre-cutover claudetools.old clone (md5-verified):
- docs/network/network-optimization-master-plan.md
- docs/network/phase1-voice-qos-design.md
- reports/2026-06-18-voice-quality-diagnostic.md
- session-logs/2026-06/2026-06-18-howard-cascades-rf-voice-optimization-plan.md
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The push handlers set the button to 'skipped' then immediately ran search(), which
re-rendered the inspector and wiped the text — so a skipped publish flashed and
vanished (looked like nothing happened). Replace with persistent toasts that state the
outcome explicitly: Published / already up-to-date / Push failed / and for a skip,
'<model> isn't renderable yet, so nothing was sent.' Only refresh the row on an actual
publish so the message isn't clobbered. Same for the multi-select Re-push summary.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Results pane now fills the full width on load (the empty ~500px inspector no longer
reserves space). Inspector column expands when a record opens; Esc closes it back to
full-width results. Same behavior at the 1180px breakpoint.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
coord.py 'lock release' takes the lock ID; the documented path form no-ops
and strands the lock until TTL. Capture the lock ID at claim (5.0), release
it in Phase 6. Recurring friction (errorlog 2x).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
New MSP tool — single-assessor consult intake. Stack: PHP + MySQL on the IX cPanel
host, gated by Cloudflare Access (only mike@azcomputerguru.com; app re-checks the
Cf-Access-Authenticated-User-Email header).
- app/questions.json — risk-ordered question framework (9 sections); each field tagged
source=syncro/rmm/scan/ask so the consult asks only what a human knows and the
post-meeting scan fills the technical reality.
- app/index.php — wizard UI: Syncro phone lookup -> prefill, section rail with live
progress, importance-colored question cards, in-meeting 365/Google consent links,
review + export.
- app/api.php — Syncro lookup-by-phone, save/load/list, consent-URL generation (reuses
the read-only Security Investigator app bfbc12a4-...), HTML export.
- app/schema.sql, config.sample.php, DEPLOY.md, README.md.
Consent links let the client approve read-only 365/Google access during the consult so
the audit scan runs afterward. Read-only by design (reads Syncro, generates consent; no
tenant writes).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Probed each of the 1,578 unuploaded PASS serials across the 54 validated DSCA33/45
models against the Hoffman API (stale inventory -> can't trust api_uploaded_at as
"absent"). 1,452 were absent (404), 126 already live. Pushed ONLY the absent set:
created=1452 updated=0 unchanged=0 errors=0 — zero overwrites of pristine originals,
the handoff's hard requirement.
Tools: publish-dsca3345-gap.js (absent-only Created publish), validate-dsca3345.js.
Net: DSCA33/45 effort complete — 54/56 models live + validated; 2 rounding-boundary
holdouts and 2 no-original models (24 units) remain null.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
