mistersilver-uk

Audit and improve Fabricate's Svelte UI, Foundry window layouts, and interaction flows. Use for UX reviews, visual design proposals, accessibility checks, responsive behavior, screenshot analysis, or creating UI backlog tasks for `src/ui/`, `styles/`, and related specs.

Implement a single planned Fabricate change in the JavaScript, Svelte, and Vite codebase with assignment-approved focused checks. Use when the issue's OpenSpec delta defines the task and code, canonical spec, or test files need to change under `src/`, `openspec/specs/`, `tests/`, or related runtime files, with the driver responsible for complete validation after integration.

Plan and coordinate Fabricate work from GitHub issues into an OpenSpec change delta authored in the issue. Use when starting any non-trivial change, selecting the next unblocked issue, re-planning after scope changes, or routing work through implementer, reviewer, and docs stages without writing production code.

Perform an independent review of Fabricate changes for correctness, regression risk, test quality, and Foundry V13 compatibility. Use after implementation is complete, when the user asks for a review, or before docs and issue closure.

Verify Fabricate's integration with the Foundry VTT API and lifecycle. Use at design time and in implementation review whenever a change calls Foundry APIs or hooks into Foundry's lifecycle; it researches Foundry sources, then the official API docs, then community discussions to confirm the real shape and behaviour of Foundry and keep Fabricate's integration seamless.

Audit and refine Fabricate's crafting domain model, ubiquitous language, and spec-to-code alignment. Use for naming decisions, DDD modelling, DOMAIN.md updates, spec or code audits, research into other crafting systems, or backlog tasks about domain fidelity.

Scan Fabricate for likely defects, edge cases, testing gaps, and unreliable UI behavior, then record actionable issues. Use for bug hunts, reliability audits, regression-risk analysis, or creating GitHub issues instead of directly implementing fixes.

Research Fabricate's competitive landscape, product differentiation, and monetization options, then maintain `COMPETITIVE_ANALYSIS.md`. Use for market-fit analysis, rival module comparisons, crafting UX research, pricing research, or monetization strategy updates.

Use to independently review a proposed plan before implementation in Picobooks. Applies when an agent has produced a plan for repository changes and must decide whether there is blocking feedback on intent, scope, constraints, tests, reviewability, or required skills before implementation begins.

Use to independently review completed repository changes before PR creation or update in Picobooks. Applies after implementation and before release to check diffs, tests, constraints, skill adherence, reviewability, and whether blocking feedback remains.

Use when planning/designing, implementing, or reviewing domain code, especially Java/Spring domain and application changes in Picobooks. Applies Domain-Driven Design modeling, Elegant Objects object design, and testability heuristics to keep domain objects behavior-rich, bounded, and easy to unit test.

Use for this repository when planning, reviewing plans, implementing, reviewing implementations, releasing, creating or editing PRs, writing commits, or responding to review comments. Enforces staged plan-review-implement-review delivery, branch-based delivery, Conventional Commit messages and PR titles, short commit bodies, non-draft PRs, local gates, and human review before merge.