pynecore-practicelisted
Install: claude install-skill theonetrade/pynecore-skills
# Pine-first: `content/<name>.pine` → `scripts/<name>.py` → `pyne run`
The user writes strategies in Pine Script and runs them on PyneCore. The Pine source is the
only file edited by hand and **stays in `content/` after generation, for the life of the
script**; the Pyne script is generated from it by the local `pine2pyne` (install and usage
guide: `references/pine2pyne.md`), carries the same stem, and is what `pyne run` executes.
Every development iteration is the same three moves: change the `.pine`, regenerate the
`.py`, run.
Every statement below was exercised on a workdir with the demo data (PyneCore 6.12.1,
pine2pyne 0.2.11): a strategy, an indicator and a library, converted, run, edited and
re-converted.
```
<workdir>/
├── .pyne
├── content/ # Pine sources: the source of truth, edited by hand
│ ├── ema_atr.pine # strategy("...") or indicator("...")
│ └── lib/mine/Helpers/v1.pine # library("Helpers"), imported as `import mine/Helpers/1 as h`
├── scripts/ # generated by pine2pyne, same stems; never edited by hand
│ ├── ema_atr.py # header: "compiled by pine2pyne <version>"
│ ├── ema_atr.toml # inputs and settings, written by the first run (see rule 4)
│ └── lib/mine/Helpers/v1.py # imported by the script as `lib.mine.Helpers.v1`
├── data/ # candle files (demo.ohlcv + demo.toml, downloads)
└── output/ # ema_atr.