Agent Skills: Implement Lorcana Card Batch

Implement a bounded batch of Lorcana cards with isolated ownership, focused tests, and shared-engine backpressure. Use when the user supplies a card list or asks to process missing implementations.

UncategorizedID: TheCardGoat/tcg-engines/source-command-implement-batch

Install this agent skill to your local

pnpm dlx add-skill https://github.com/TheCardGoat/tcg-engines/tree/HEAD/submodules/lorcana/.agents/skills/source-command-implement-batch

Skill Files

Browse the full folder contents for source-command-implement-batch.

Download Skill

Loading file tree…

submodules/lorcana/.agents/skills/source-command-implement-batch/SKILL.md

Skill Metadata

Name
source-command-implement-batch
Description
Implement a bounded batch of Lorcana cards with isolated ownership, focused tests, and shared-engine backpressure. Use when the user supplies a card list or asks to process missing implementations.

Implement Lorcana Card Batch

Use lorcana-cards as the behavior owner and lorcana-find-card to resolve each card. There is no maintained source-file queue; never delete source rows as a completion signal.

Workflow

  1. Build a bounded queue from the user's list or an explicit missingImplementation: true search.
  2. Record each card's definition and test paths before assigning work.
  3. Parallelize only cards with disjoint definition and test files. Keep shared engine, types, exports, and generator files under one owner.
  4. Require one focused behavior test and one structured result per card: implemented, already-done, skipped, or blocked.
  5. If two cards expose the same engine gap, stop the batch and fix that shared primitive once before continuing.
  6. Review the combined diff, run affected card tests, then run the package check once.

Do not remove source records, stage broad paths, or let concurrent workers edit the same shared file.