Re-ran benchmark.py on GURU-BEAST-ROG against the post-overhaul code
(co-host profile, batched Whisper int8_float16, revised Q&A extractor).
Results vs 5070 Ti baseline:
- Diarization: 209.7x -> 338.1x (+61.2%)
- Transcription: 63.8x -> 94.8x (+48.6%)
- Q&A pairs: 9 vs 10 (within run-to-run noise; structural correctness matches:
2014 = 0 callers, 2016 = 2 WiFi caller pairs)
Setup change: BENCH_SETUP.md now lists ffmpeg as a Step-2 prereq
(winget install Gyan.FFmpeg). Was missing on this machine and the pipeline
fails silently at the first diarize call without ffprobe.
Code change: benchmark.py BASELINE_RTF updated 149.5 -> 209.7 to reflect
the 5070 Ti's post-overhaul measurement (e9ac607).
Data: 6 test episode transcripts and diarizations regenerated under the
new code path (batched Whisper output + co-host-aware speaker_map).
Correction memory: voice-profiles/tom/ directory + 5070 Ti session log
fabricated a co-host named "Tom" — Mike confirms no such person exists on
the show. The audio profile is real and the diarization separation is
sound, but the human identity attached to it is wrong. Saved under
.claude/memory/radio_show_no_cohost_named_tom.md pending Mike providing
the correct name for rename.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Memory entry prompts Mac session to run scripts/install-hooks.sh
before any GuruRMM work. Syncs via Gitea on next pull.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- .claude/identity.json (gitignored, per-machine) identifies who's at the keyboard
- .claude/users.json (tracked) registers known team members + roles + machines
- CLAUDE.md: on first sync, Claude asks "Mike or Howard?" and creates identity.json
- Session logs must include User section for attribution
- Git commits use per-user name/email (shared Gitea push account)
- Howard Enos (tech, full trust) added as second team member
- Memory entry created for Howard
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
New integration with TickTick API for project/task management:
- OAuth 2.0 auth flow (mcp-servers/ticktick/ticktick_auth.py)
- MCP server with 9 tools for Claude Code (ticktick_mcp.py)
- FastAPI service with SOPS vault credentials (api/services/ticktick_service.py)
- JWT-protected REST router at /api/ticktick/ (api/routers/ticktick.py)
- Credentials stored in SOPS vault (services/ticktick.sops.yaml)
Dev project tracking (hybrid TickTick + DB):
- New dev_projects table migration (14 columns, status index)
- TickTick "Dev Projects" list for mobile visibility
- First project seeded: TickTick Integration (linked both sides)
Security: .tokens.json gitignored, token file permissions restricted,
HTML-escaped OAuth callback, SOPS vault (not env vars) for secrets.
Also: Installed Tailscale on ACG-5070 for office network access.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
