diff --git a/.agents/rules/frontend-design.md b/.agents/rules/frontend-design.md new file mode 100644 index 0000000..f02c13a --- /dev/null +++ b/.agents/rules/frontend-design.md @@ -0,0 +1,16 @@ +# Frontend Design Rules + +## Routing Rule +ROUTING RULE: You must structure the Next.js App Router strictly by business domains. The allowed directories are exactly: /app/(dashboard)/jobs, /app/(dashboard)/candidates, /app/(dashboard)/interviews, and /app/(dashboard)/workflows. + +## Styling Whitelist +STYLING WHITELIST: You are restricted to a neutral, professional Human Resources design system. You must ONLY use the following Tailwind CSS tokens: +- **Backgrounds:** `bg-slate-50` for main layouts, `bg-white` for cards/surfaces. +- **Text:** `text-slate-900` for headings, `text-slate-600` for body text, `text-slate-500` for secondary labels. +- **Primary Actions (Buttons/Links):** `bg-blue-600` for backgrounds, `hover:bg-blue-700` for hover states, `text-white` for text. +- **Borders & Dividers:** `border-slate-200`. +- **Border Radius:** strictly `rounded-md` or `rounded-lg`. Do not use full rounding unless it is an avatar. +- **Shadows:** strictly `shadow-sm` for cards, `shadow-md` for modals/dropdowns. + +## UI Component Rule +UI COMPONENT RULE: Build and reuse standard components (e.g., ``, ``, ``). Do not invent new visual styles outside of the allowed tokens. diff --git a/.agents/rules/quality-and-ai.md b/.agents/rules/quality-and-ai.md new file mode 100644 index 0000000..c8a157a --- /dev/null +++ b/.agents/rules/quality-and-ai.md @@ -0,0 +1,7 @@ +# Quality and AI Rules + +## Git Hooks Rule +GIT HOOKS RULE: The repository must enforce quality through local Git hooks. You are required to configure a pre-commit hook for formatting, a commit-msg hook for commit conventions, and a pre-push hook for testing. + +## AI Output Schema Rule +AI OUTPUT SCHEMA RULE: Any integration with an LLM for parsing or evaluating candidates must output a deterministic JSON structure. You are restricted to generating JSON that contains EXACTLY these four keys: `summary`, `classification`, `suggestions`, and `riskLevel`.