chore: improve automatic rule detection for LLMs (#5588)

Author: therealemjy

.claude/CLAUDE.md

@@ -1,6 +1,6 @@
-# Venus Protocol Interface - Claude Rules
+# Venus Protocol Interface - Rules
 
-Before generating or modifying code in this repository, Claude must read:
+Before generating or modifying code in this repository, you must read:
 
 1. `.claude/references/design-system.md`
 2. `.claude/references/api-patterns.md`

.claude/references/code-quality.md

@@ -187,8 +187,12 @@ Regenerate using project scripts (`yarn generate-*`, `yarn generate`) instead.
 ## Testing stack and expectations
 
 - Unit/integration tests use Vitest.
+- When writing tests within a workspace, first read its Vite config and setupTests files (if any) to
+  ensure the tests you write are aligned with them.
 - React rendering tests use Testing Library.
 - Test filenames in this repo use `*.spec.ts` / `*.spec.tsx`.
+- Tests should be run using the Node version specified in the repo `.nvmrc`.
+- Only mock components when absolutely necessary.
 
 ### Practical testing guidance
 

AGENTS.md

@@ -0,0 +1,3 @@
+# Venus Protocol Interface - Agent Bootstrap
+
+Before generating, editing, or reviewing code in this repository, read `.claude/CLAUDE.md`.

CLAUDE.md

@@ -0,0 +1,3 @@
+# Venus Protocol Interface - Claude Bootstrap
+
+Before generating, editing, or reviewing code in this repository, read `.claude/CLAUDE.md`.

package.json

@@ -3,17 +3,13 @@
   "version": "1.0.0",
   "private": true,
   "packageManager": "yarn@4.14.1",
-  "workspaces": [
-    "packages/*",
-    "configs/*",
-    "apps/*"
-  ],
+  "workspaces": ["packages/*", "configs/*", "apps/*"],
   "scripts": {
     "start": "turbo run start",
     "generate": "turbo run generate",
     "lint": "turbo run lint",
     "lint:styles": "turbo run lint:styles",
-    "format": "biome check --apply .",
+    "format": "npx biome check --apply .",
     "test": "turbo run test --",
     "tsc": "turbo run tsc",
     "clean": "turbo run clean && rm -rf node_modules",