jyapayne/houserules
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
No description
192 commits 1 branch 0 tags 4.5 MiB
Find a file
Joey Yakimowich-Payne a159299818
 chore(sisyphus): execution plan for rule-variants epic (v2) 
Writes the step-by-step to-do list successor to the Momus-approved
design doc at .sisyphus/plans/rule-variants.md. The design doc has
the full hook-API rationale, risk analysis, and architecture sketch;
rule-variants-v2.md is the agent-executable task list.

Structure mirrors the modifier-profiles-t3 plan the repo just
shipped: 6 phases (A-F) + Final Verification Wave, 32 top-level
checkboxes, parallel-execution map, T3-learned tactics section for
the incoming agent (tool-cap fragility, commit discipline,
v3/v4 mirror pitfalls), and explicit 'when to stop mid-plan'
branching so an early-exit delivers a coherent subset.

Scope (recap from design doc):
  - Phase A: 4 new PresetDef hooks (getRoyalPieces,
    filterLegalMoves, shouldAdvanceTurn, overridePieceMoves) +
    HalfMovesThisTurn game fact.
  - Phase B: 4 Tier-1 presets closing starting-layouts deferrals
    (knightmate-rules, double-move, monster-rules,
    first-promotion-wins).
  - Phase C: 3 royal-variant Tier-2 presets (coregal, dual-king,
    weak-dual-king) + dual-classic layout.
  - Phase D: 3 objective-variant Tier-2 presets (suicide-chess,
    capture-all, extinction-chess).
  - Phase E: 4 movement-variant Tier-2 presets (berolina-pawns ×2,
    bouncing-pieces ×2).
  - Phase F: lobby integration (incompatibleWith audit,
    suggestedPresets wiring, drawer category grouping), rule-
    variant e2e, docs.
  - Final Wave: 4 reviewers (oracle/deep/unspecified-high).

Explicit deferrals: royalty-transfer (T3 in design doc; needs
mid-game state toggle), PGN notation changes, ELO/ranking,
non-8×8 variants, multi-royal check-banner UX (follow-up).

Each task carries 'what to do', file anchors, verification
criteria, parallelization map, recommended agent profile, and
commit message template — designed so an agent can pick up task
X.Y without re-reading the whole plan.
2026-04-20 18:15:36 -06:00
.github/workflows chore(root): scaffold monorepo — Phase 0 complete 2026-04-16 13:32:21 -06:00
.sisyphus chore(sisyphus): execution plan for rule-variants epic (v2) 2026-04-20 18:15:36 -06:00
docs docs(adr): T4 scripted modifiers forward-design 2026-04-19 21:09:52 -06:00
packages fix(modifiers): close client schema/server schema drift + add parity test (Q4.2) 2026-04-20 17:53:22 -06:00
scripts feat(rete): add replay engine + state-hash determinism verifier (P3.3) 2026-04-16 15:25:13 -06:00
.gitignore chore: ignore .org.chromium.Chromium.* runtime files 2026-04-19 10:20:42 -06:00
eslint.config.js feat(engine): damage-resistance modifier descriptor 2026-04-18 22:22:31 -06:00
lefthook.yml chore(root): scaffold monorepo — Phase 0 complete 2026-04-16 13:32:21 -06:00
LICENSE chore(root): scaffold monorepo — Phase 0 complete 2026-04-16 13:32:21 -06:00
package.json feat(rete): add replay engine + state-hash determinism verifier (P3.3) 2026-04-16 15:25:13 -06:00
playwright.config.ts fix(rete): inject clock into EventLog; use tsc for DTS; fix cycle.test.ts private access; add Playwright worker limit 2026-04-16 18:25:49 -06:00
README.md chore(root): scaffold monorepo — Phase 0 complete 2026-04-16 13:32:21 -06:00
tsconfig.base.json chore(root): scaffold monorepo — Phase 0 complete 2026-04-16 13:32:21 -06:00
tsconfig.json chore(root): scaffold monorepo — Phase 0 complete 2026-04-16 13:32:21 -06:00
vitest.workspace.ts chore(root): scaffold monorepo — Phase 0 complete 2026-04-16 13:32:21 -06:00

README.md

@paratype

A Doorenbos-style Rete II rules engine for TypeScript games, with an authoritative WebSocket chess demo.

Packages

Docs

Getting Started

bun install && bun run check