chart/README.md
Chris Amow 8e50d5cbc2 Add implementation plan for /ES multi-timeframe confluence chart
Planning-only commit: no application code yet.

The plan specifies a realtime /ES chart that derives moving averages and
trendlines across multiple timeframes, projects them onto one chart in a
shared (time, price) plane, and alerts when levels from different
timeframes converge.

Key findings that shaped it, all verified against source rather than
assumed:

- Schwab streams realtime futures fine (CHART_FUTURES, LEVEL_ONE_FUTURES)
  but provides no futures price *history* at all. An account does not
  change this; it is an API-surface limit.
- Yahoo's chart endpoint needs no key and has exactly what Schwab lacks:
  ~730d of hourly ES=F (~750 sessions), enough to warm a 200DMA from
  startup. So it serves as both the no-keys dev source and the history
  seeder, behind one MarketDataSource protocol.
- Yahoo anchors daily bars to midnight ET while the CME session runs
  18:00-17:00 ET, so daily bars are built from hourly using our own
  session rules instead.
- Lightweight Charts v5 replaced addCandlestickSeries() with
  addSeries(CandlestickSeries, ...); most tutorials online are v4.

Build order defers judgment-heavy work: moving averages first (fully
deterministic), then confluence scoring, then hand-drawn trendlines.
Automatic trendline detection comes last, tuned against the hand-drawn
lines as ground truth.

Includes a real trimmed Yahoo response as a test fixture; it contains a
null in the OHLC arrays, which is the parsing case that needs handling.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-08-09 20:31:40 -05:00

66 lines
2.2 KiB
Markdown

# chart
FastAPI backend + Vue 3 (from CDN, no build step) served at
<https://chart.amow.com>.
Currently a placeholder: the frontend calls `/api/hello` and prints the JSON.
**Where this is going:** a realtime `/ES` chart that derives trendlines and moving
averages across many timeframes and alerts when they converge. The full spec lives in
[`docs/IMPLEMENTATION_PLAN.md`](docs/IMPLEMENTATION_PLAN.md) — read it before writing
code; it records decisions and verified API facts that are expensive to rediscover.
## Local development
```bash
docker compose up --build
```
Then open <http://localhost:8000>. Override the host port with
`PORT=8010 docker compose up` if 8000 is busy — on the VPS itself it always is,
that's the Coolify UI. The source tree is bind-mounted and uvicorn
runs with `--reload`, so edits to `main.py` or `static/` take effect without a
rebuild. Rebuild only when `requirements.txt` changes.
Without Docker:
```bash
python3 -m venv .venv && . .venv/bin/activate
pip install -r requirements.txt
uvicorn main:app --reload
```
## Layout
| Path | Purpose |
|---|---|
| `main.py` | FastAPI app — JSON under `/api`, serves the SPA at `/` |
| `static/` | `index.html`, `app.js`, `style.css` — Vue 3 loaded from unpkg |
| `requirements.txt` | Python deps |
| `Procfile` | Start command; **nixpacks needs this** or the deploy has nothing to run |
| `Dockerfile.dev`, `docker-compose.yml` | Local dev only — production does not use them |
## Deployment
Push to `main` → Forgejo webhook → Coolify rebuilds with nixpacks → live in
roughly 25 seconds.
```bash
git push
```
Two things to know before changing the deploy config:
- The Coolify domain is registered as `https://chart.amow.com:8000`. The port
suffix is how Coolify decides the Traefik target port; drop it and every
request 502s while the container looks perfectly healthy. It does not appear
in the public URL.
- Ignore a 502 in the first minute after a deploy — that's the rolling container
swap, and it clears on its own.
Manual redeploy:
```bash
curl -X POST -H "Authorization: Bearer $(cat ~/.coolify-token)" \
"http://127.0.0.1:8000/api/v1/deploy?uuid=dgvch0xqv8uvjfor7dl8bwl9&force=true"
```