By name: orchestra description: "Round-robin multi-LLM consultation tuned for InTheWake PWA development. GPT proposes, Gemini refines, Grok challenges — full-context debate with wheat/chaff justifications, low-hanging fruit, and PWA checklist." version: 1.0.0
The Orchestra — InTheWake PWA Edition
Ask the orchestra. Let them debate. Build what survives.
Usage
cd /home/user/ken/orchestrator && python3 orchestra.py cruising "task description"
How It Works
Round-robin debate where each model sees EVERYTHING from prior rounds — proposals, verdicts, justifications, rejections. Nothing filtered. What one model calls chaff, the next might rescue.
The Rounds
Round 1: GPT — Proposes structure, approach, innovations. Each proposal has CONFIDENCE + JUSTIFICATION.
Round 2: Gemini — Sees GPT's full response. For each proposal: WHEAT | CHAFF | WHEAT_WITH_REFINEMENT (with justification). Adds own proposals. Identifies low-hanging fruit.
Round 3: Grok — Sees both GPT + Gemini. Verdicts on EVERY prior proposal. Specifically rescues dismissed ideas worth saving. Adds own proposals. Identifies blind spots + low-hanging fruit.
Blind Spot Check — Grok reviews the emerging synthesis: "What are we all still missing?"
Claude Synthesizes — Full debate chain → attributed plan → encode to memory.
Why Full Context
Each model provides justifications for wheat/chaff verdicts. The next model can disagree with the reasoning, not just the conclusion. An idea rejected in Round 2 can be rescued in Round 3 because the justification was visible.
PWA Checklist
After every orchestra run, verify these PWA concerns:
| Concern | Check | File |
|---|---|---|
| Service Worker | Does sw.js need updating for new pages/assets? | sw.js |
| Cache Strategy | Static assets: cache-first. Dynamic data: network-first. | sw.js |
| Manifest | App name, icons, theme color, start URL current? | manifest.json, manifest.webmanifest |
| Offline | New pages fallback to offline.html? | offline.html |
| Precache | New files added to precache manifest? | precache-manifest.json |
| Prefetch Images | New images in prefetch list? | prefetch-images.json |
| Performance | First load <200KB? LCP <2.5s? | Lighthouse |
| Sitemap | New pages listed? | sitemap.xml |
| AI Discovery | llms.txt current with new content? | llms.txt |
| Schema | POI/port data validates against schema? | schema/*.json |
Cost
Typical: $0.03–$0.08 per full orchestra run. Report shows per-round breakdown.
Integration
- cognitive-memory — recalls cruising memories before Round 1, encodes decisions after
- deployment-validator — run after orchestra to verify artifacts
- webapp-testing — run after orchestra if tools were modified
- content-freshness — orchestra may identify stale content to update
- seasonal-content-planner — orchestra informed by seasonal priorities
Soli Deo Gloria — Let the orchestra play. Build what serves the traveler.