Skip to content

AGENT_CONTEXT.md

Latest commit

 

History

History
37 lines (30 loc) · 1.65 KB

AGENT_CONTEXT.md

File metadata and controls

37 lines (30 loc) · 1.65 KB

Agent Continuation Context

Last updated: 2026-04-14 (UTC) Branch: work

What is stable now

  • Full simulation runs (python main.py).
  • Projectile API is integrated in engine/full_simulation.py.
  • Season loader supports listing/loading and basic schema checks.
  • Initial unit tests exist under tests/.
  • Trait activation pipeline now exists with initial start-of-battle effects for Ace and Undead.
  • Trait multipliers are now data-driven from data/trait_effects.py with explicit placeholder provenance.

Simplicity assessment

  • Current design is mostly not over-engineered for this stage.
  • The only risk area is scope creep (shop/draft/multi-round) before core combat fidelity is complete.
  • Keep a strict rule: only build abstractions that are used by running simulation code and covered by tests.

Known gaps

  1. AbilityEngine is still a placeholder.
  2. Trait logic is only partially implemented (currently only Ace and Undead).
  3. Most troop stats still rely on defaults because data/troops.py lacks complete numeric fields.
  4. Season JSON contains deltas but not complete baseline numeric data.

Suggested next execution order

  1. Implement engine/traits.py and integrate at battle start.
  2. Fill troop numeric stats and remove implicit defaults over time.
  3. Add ability handlers for a small subset of troops, then expand.
  4. Add integration tests around complete battle determinism.
  5. Defer non-core systems (shop/draft/rounds) until steps 1-4 are stable.

Validation baseline

  • python -m unittest discover -s tests -v
  • python main.py

Decision log

  • See DECISIONS.md for current beliefs, counterarguments, and architecture self-review.