dodoxandClaude Sonnet 5
7a538cb300
- puga/: cached FIO + PRUNplanner clients, market view with order-book walk, econ formulas ported from PRUNplanner (tested against its suite and live FIO), saturation model v1 (reviewed by Opus) - tools/: scan (depth-aware), price, book, chain, state sync, plan_push (dry run default, [PuGa]-prefixed plans only), legacy prun_scan/prun_cxarb - docs/: mechanics (PRUNplanner is source of truth), roadmap, decisions, saturation design, archived handoff - secrets stay in .env (gitignored); ref/ holds PRUNplanner source (ignored) Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
1.7 KiB
1.7 KiB
Decisions log
2026-09-18 Project plan and working agreements
- Goal: agent-first toolkit; Dominik asks, the agent answers by running tools against live data. Not an end-user app.
- PRUNplanner source (
ref/) is the source of truth for mechanics; conflicts with the old handoff resolved in its favour (docs/mechanics.md). - Scope: whole universe, default region Antares (CX AI1). Python in
.venv, secrets in.env, git branchmaster. No commits without Dominik asking. - Every reported opportunity must include market saturation (buildings the market can absorb, fill-price ROI, absorbable profit/day). Reason: PRUNplanner's ROI Overview ranks recipes with no depth awareness (e.g. 0.25 day ROI on recipes 2 buildings saturate).
- PRUNplanner integration: API key auth (
Authorization: Api-Key <key>); plans can be created/updated via/planning/plan/. Writes only create/update plans prefixed[PuGa]; never touch Dominik's other plans; show payload and get a yes before each write.
Model and review policy
- Default: Sonnet 5 does the building (API plumbing, CLIs, ports, tests).
- Bigger model (Agent tool
model: opusorfable, fresh context, brief with files and question) is allowed and expected at these points:- Saturation model design, before building it, and again on real output.
- Econ formula port (efficiency stack, workforce, hab LP, extraction) checked against PRUNplanner numbers.
plan_pushpayload handling and the own-plans-only guard.- Big-money recommendations (e.g. BHP go/no-go, roughly 100k AIC): independent second opinion on the numbers.
- If a bigger-model spawn is refused, tell Dominik and ask him to run the review.
- Routine tool building is not reviewed by a bigger model.