luke/ai-podcast
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
9e37fbf124213b67fe05c2b0e7f1bc84a0c55ad8
ai-podcast/CLAUDE.md
T
luke 9e37fbf124 Phase 5B commit 6: delete CallerBackground dataclass, relationship_context, update docs 
Final cleanup pass for the caller generation redesign:
- Delete CallerBackground dataclass + all isinstance() checks
- Delete orphaned _build_relationship_context and session.relationship_context
- Delete Session.get_caller_model (caller_dialog category routes to haiku-4.5)
- Delete dead _get_show_energy (CallRecord no longer tracks energy_level)
- Delete unused tone_streak field
- Drop topic_category/emotional_state/energy_level from CallRecord
- Simplify caller property, _find_thematic_match, enrich, promotion paths to slim dict
- Drop broken emotional_state=/energy_level= kwargs from generate_speech calls
- _load_checkpoint drops pre-slim schema backgrounds; startup re-pregens if empty
- Rewrite CLAUDE.md "Caller Generation System" section for the new two-stage architecture

35 tests passing. backend/main.py: 4981 -> 4810 lines.
2026-04-05 13:56:08 -06:00

7.4 KiB
Raw Blame History

AI Podcast - Project Instructions

Git Remote (Gitea)

  • Repo: git@gitea-nas:luke/ai-podcast.git
  • Web: http://mmgnas:3000/luke/ai-podcast
  • SSH Host: gitea-nas (configured in ~/.ssh/config)
    • HostName: mmgnas (use mmgnas-10g if wired connection issues)
    • Port: 2222
    • User: git
    • IdentityFile: ~/.ssh/gitea_mmgnas

NAS Access

  • Hostname: mmgnas (wireless) or mmgnas-10g (wired/10G)
  • SSH Port: 8001
  • User: luke
  • Docker path: /share/CACHEDEV1_DATA/.qpkg/container-station/bin/docker

Castopod (Podcast Publishing)

  • URL: https://podcast.macneilmediagroup.com
  • Podcast handle: @LukeAtTheRoost
  • API Auth: Basic auth (credentials in .env: CASTOPOD_USERNAME, CASTOPOD_PASSWORD)
  • Container: castopod-castopod-1
  • Database: castopod-mariadb-1 (user: castopod, db: castopod)

Running the App

# Start backend — ALWAYS use --reload-dir to avoid CPU thrashing from file watchers
python -m uvicorn backend.main:app --reload --reload-dir backend --host 0.0.0.0 --port 8000

# Or use run.sh
./run.sh

Publishing Episodes

python publish_episode.py ~/Desktop/episode.mp3

Environment Variables

Required in .env:

  • OPENROUTER_API_KEY
  • ELEVENLABS_API_KEY (optional)
  • INWORLD_API_KEY (for Inworld TTS)

Post-Production Pipeline (added Feb 2026)

  • Branch: feature/real-callers — all current work is here, pushed to gitea
  • Stem Recorder (backend/services/stem_recorder.py): Records 5 WAV stems (host, caller, music, sfx, ads) during live shows. Uses lock-free deque architecture — audio callbacks just append to deques, a background writer thread drains to disk. write() for continuous streams (host mic, music, ads), write_sporadic() for burst sources (caller TTS, SFX) with time-aligned silence padding.
  • Audio hooks in backend/services/audio.py: 7 tap points guarded by if self.stem_recorder:. Persistent mic stream (start_stem_mic/stop_stem_mic) runs during recording to capture host voice continuously, not just during push-to-talk.
  • API endpoints: POST /api/recording/start, POST /api/recording/stop (auto-runs postprod in background thread), POST /api/recording/process
  • Frontend: REC button in header with red pulse animation when recording
  • Post-prod script (postprod.py): 6-step pipeline — load stems → gap removal → voice compression (ffmpeg acompressor) → music ducking → stereo mix → EBU R128 loudness normalization to -16 LUFS. All steps skippable via CLI flags.
  • Known issues resolved: Lock-free recorder (old version used threading.Lock in audio callbacks causing crashes), scipy.signal.resample replaced with nearest-neighbor (was producing artifacts on small chunks), sys import bug in auto-postprod, host mic not captured without persistent stream

LLM Settings

  • _pick_response_budget() in main.py controls caller dialog token limits (150-450 tokens). MiniMax respects limits strictly — if responses seem short, check these values.
  • Default max_tokens in llm.py is 300 (for non-caller uses)
  • Grok (x-ai/grok-4-fast) works well for natural dialog; MiniMax tends toward terse responses
  • generate_with_tools() in llm.py supports OpenRouter function calling for the intern feature

Caller Generation System

  • Two-stage pipeline: (1) batch identity pregen via Sonnet 4.6 at session start, (2) live dialog via Haiku 4.5 per turn. Cost ~$1/show.
  • Slim caller dict: Populated once at Session._pregenerate_backgrounds() via caller_gen.generate_batch(). Keys: name, age, voice, location, identity, situation, reason_calling, opening_line, secret_want, specific_details, emotional_register. Stored in session.caller_backgrounds[caller_key].
  • Dialog model: Always Haiku 4.5 via the caller_dialog category in config.category_models. No per-caller model routing — deleted in Phase 5B.
  • Prompt builder: get_caller_prompt(caller) in main.py builds the slim system prompt from the dict; see tests/test_caller_prompt.py for the contract.
  • Regulars: backend/services/regulars_v2.py loads lore from Obsidian markdown files for named recurring callers (e.g. Silas). The batch prompt optionally includes 2-3 active regulars per session.
  • Inter-caller awareness: get_show_history() scores previous callers by keyword overlap with the current caller's situation/reason_calling. Reaction frequency scales with match strength (60%/35%/15%).
  • Caller memory: Returning callers auto-promote from first-timers at ~5% probability after 8+ exchanges. RegularCallerService tracks summaries, relationships, arc state.
  • Call quality signals: _assess_call_quality() captures exchange count, response length, host engagement, caller depth, natural ending.

Devon (Intern Character)

  • Service: backend/services/intern.py — persistent show character, not a caller
  • Personality: 23-year-old NMSU grad, eager, slightly incompetent, gets yelled at. Voice: "Nate" (Inworld), no phone filter.
  • Tools: web_search (SearXNG), get_headlines, fetch_webpage, wikipedia_lookup — via generate_with_tools() function calling
  • Endpoints: POST /api/intern/ask, /interject, /monitor, GET /api/intern/suggestion, POST /api/intern/suggestion/play, /dismiss
  • Auto-monitoring: Watches conversation every 15s during calls, buffers suggestions for host approval
  • Persistence: data/intern.json stores lookup history
  • Frontend: Ask Devon input (D key), Interject button, monitor toggle, suggestion indicator with Play/Dismiss

Frontend Control Panel

  • Keyboard shortcuts: 1-0 (callers), H (hangup), W (wrap up), M (music toggle), D (ask Devon), Escape (close modals)
  • Wrap It Up: Amber button that signals callers to wind down gracefully. Reduces response budget, injects wrap-up signals, forces goodbye after 2 exchanges.
  • Caller info panel: Shows identity, situation, signature detail, secret want during active calls
  • Caller buttons: Populated from the slim caller background dicts
  • Pinned SFX: Cheer/Applause/Boo always visible, rest collapsible
  • Visual polish: Thinking pulse, call glow, compact media row, smoother transitions

Website

  • Domain: lukeattheroost.com (behind Cloudflare)
  • Analytics: Cloudflare Web Analytics (enable in Cloudflare dashboard, no code changes needed)
  • Deploy: npx wrangler pages deploy website/ --project-name=lukeattheroost --branch=main

Git Push

  • If mmgnas times out, use the 10g hostname: 
    GIT_SSH_COMMAND="ssh -o HostName=mmgnas-10g -p 2222 -i ~/.ssh/gitea_mmgnas" git push origin main

Hetzner VPS

  • IP: 46.225.164.41
  • SSH: ssh root@46.225.164.41 (uses default key ~/.ssh/id_rsa)
  • Specs: 2 CPU, 4GB RAM, 38GB disk (~33GB free)
  • Mail: docker-mailserver at /opt/mailserver/
  • Manage accounts: docker exec mailserver setup email add/del/list
  • Available for future services — has headroom for lightweight containers. Not suitable for storage-heavy services (e.g. Castopod with daily episodes) without a disk upgrade or attached volume.

Podcast Workflow

  • Publishing pipeline: episodes go through Castopod, CDN, website, YouTube, and social
  • Always check Python venv is active and packages are installed before running publish scripts
  • Episode numbering must be verified against existing episodes

Episodes Published

  • Episode 6 published 2026-02-08 (podcast6.mp3, ~31 min)
Reference in New Issue View Git Blame Copy Permalink