# PuGa: Prosperous Universe advisory toolkit

Agent-first toolkit for advising Dominik (username dodox, company GBI, faction Antares, home CX AI1). He asks questions; you answer by running tools in `tools/` against live data. Read this file first, then `docs/mechanics.md` and `state/company.yaml`.

## How to work with Dominik
- Expert user. Numbers, tables, ROI per day. Short answers, no fluff, no moralising. Semicolons over dashes; no em dashes.
- He plays on browser APEX (free licence) and sends screenshots; read them carefully, they override assumptions.
- Never guess APEX commands or numbers. Pull data (tools, FIO, PRUNplanner) or search. Command list: `docs/apex.md` (when written) and `docs/handoff-2026-09-18.md` §4.
- Label every number: live (tool output), state (company.yaml), or estimate; say which in-game command would confirm an estimate.
- Prices move; re-run tools before quoting. Snapshot prices in the handoff doc are stale.

## Source-of-truth rules
1. **PRUNplanner code wins** over our own docs for game mechanics (`ref/frontend/src/features/planning/calculations/`, `ref/backend/`). If `docs/handoff-2026-09-18.md` conflicts with it, the handoff is wrong; fix `docs/mechanics.md`.
2. In-game numbers Dominik reports beat both, especially for planet resource factors.
3. `ref/` is a gitignored copy of the PRUNplanner repos; refresh with `tools/refresh_refs.sh`.

## Setup
- Python venv at `.venv` (`.venv/bin/python`, `.venv/bin/pip`). Deps in `requirements.txt`.
- Secrets in `.env` (gitignored; template `.env.example`): FIO REST key, FIO API key, PRUNplanner key. Never print or commit them; do not ask Dominik to paste keys into chat.
- Scope: whole universe supported; default region Antares (exchange AI1, `--cx` to change).

## Layout
- `CLAUDE.md` this file; `docs/` mechanics, refs, roadmap, decisions, archived handoff.
- `state/company.yaml` company state (bases, buildings, ships); update when Dominik reports changes.
- `puga/` shared library (data layer, market, econ, world); `tools/` CLIs; `data/cache/` fetched data; `tests/`.
- `ref/` PRUNplanner source (read-only reference).

## Tools (current)
- `tools/state.py sync|show` refreshes `state/company.yaml` from live FIO (cash, permits, buildings, real production efficiency, storage, ships). Run it before answering anything about his current position.
- `tools/plan_push.py` builds/validates PRUNplanner plans from `plans/*.yaml` (dry run by default; `--apply` only after Dominik says yes; only `[PuGa]`-named plans are created/updated; `list` shows his plans). Specs live in `plans/`.
- `tools/chain.py TICKER` make-vs-buy cost tree plus sourcing depth of inputs.
- `tools/book.py` order-book ladder (price levels, cumulative units) for one material at one CX.
- `tools/price.py` prices across CX with VWAP, daily volume, and order-book fill price for a quantity (`--qty`). Library: `puga/market.py`, `puga/fio.py`, `puga/prunplanner.py`.
- `tools/scan.py` DEPTH-AWARE recipe scan (use this): N* buildings the market absorbs, ROI at N=1 and N*, patient prices, ask-walked inputs. Flags: --cx --tier --cogc --skip <tier> --hq --budget --min-n --sort roi|total|roi1 --show-thin. Library: `puga/saturation.py`.
- `tools/prun_scan.py` legacy single-step scan at top-of-book prices; overstates thin markets. Baseline only.
- `tools/prun_cxarb.py` inter-exchange arbitrage per material. Legacy; overstated profit 7 to 11x on thin routes; `tools/arb.py` replacement is on the roadmap.

- `puga/econ.py` pure formulas ported from PRUNplanner: efficiency stack, workforce satisfaction/consumption, extraction, production I/O, hab optimizer. `tests/test_econ.py` holds the reference values.

## Model policy
Sonnet builds; spawn a bigger model (Agent `model: opus|fable`) for the review points in `docs/decisions.md`. Roadmap: `docs/roadmap.md`.

## Key analytic rule: depth matters
PRUNplanner's ROI Overview ranks recipes as if the market absorbs unlimited output (e.g. 0.25 day ROI on a recipe whose whole market fits in 2 buildings). Every opportunity we report must include **saturation**: max buildings the market can absorb on both the output side (demand, traded volume, order-book walk) and the input side (supply). Report ROI at realistic fill prices, not top-of-book, and total absorbable profit/day.