Book 0 · Lesson 0.9

The God Project mental model

  • Map the six planning PDFs to monorepo directories
  • Explain the Academy-to-Team-Radar learning loop

The God Project is not one app — it’s a learning loop: each Academy lesson teaches a concept you immediately implement in Team Radar.

The six PDFs → this repo

PDFVisionMaps to
Gemini1Monorepo modelRoot layout, apps/, services/, packages/
Gemini2Team Radar diagramEvent flow: inputs → gateway → stream → dashboard
Gemini3Rust + Python, Kafka/PG/Redisservices/, infra/docker/
Gemini4JWT gateway + mTLSservices/gateway/, infra/certs/
Gemini5Chained Consumer, 1:1 portalLLM worker → analyzer → manager views
Gemini6WASM redaction, SolidJSpackages/redact-wasm/, apps/radar/

Source PDFs live in docs/ — they are planning artifacts, not runtime code.

Monorepo mental model

millipede/
├── apps/academy/     ← you are here (teach)
├── apps/radar/       ← SolidJS dashboard (ship Stage 4)
├── services/         ← Rust + Python pipeline
├── packages/         ← shared widgets, WASM, types
├── content/          ← MDX books 0–7 (mirrors Academy)
└── infra/docker/     ← Kafka + Postgres + Redis

The pipeline you’ll build

Axum Ingestionraw-dev-events
Python LLM Workerenriched-dev-events
Rust AnalyzerPostgres + Redis
SolidJS Dashboardlive metrics
  1. Ingestion (Rust Axum) accepts webhooks, publishes to raw-dev-events
  2. LLM worker (Python) enriches sentiment and risk, publishes enriched-dev-events
  3. Analyzer (Rust) writes Postgres + Redis for live dashboard
  4. Gateway (Rust) validates JWT, proxies over mTLS — no Express BFF
  5. Dashboard (SolidJS + TanStack) renders team health for managers

Work repos are reference only

See docs/work-alignment.md — maverick, maverick-ui, and ai-pentest inform patterns (eval CI, OTel, BFF deep-dive) but do not reshape this stack.

Capstone prep

Draw Team Radar on a whiteboard: inputs, Kafka topics, services, and the manager dashboard. Time yourself — 5 minutes.