Add recursive type analysis for nested sensitive data detection
Rewrote type analyzer to properly handle nested object types and
detect sensitive data at any depth.
**New Analysis Strategy:**
1. **Primitive types** (string, number, boolean):
- Check prop name for sensitive patterns
- Example: `password: string` → SENSITIVE
2. **Object types**:
- Recursively traverse all nested properties
- Check each nested key for sensitive patterns
- Example: `config.credentials.password` → SENSITIVE
3. **Function types**:
- Warn that functions are non-serializable
- Example: `onClick: () => void` → Warning
**Implementation:**
- Added `analyzePropType()` recursive function
- Checks TypeScript type flags to determine primitive vs object
- Builds full property paths (e.g., `config.credentials.password`)
- Detects sensitive data at any nesting level
**Example Detection:**
```typescript
interface Props {
config: { // object - recurse
credentials: { // object - recurse
password: string // primitive - SENSITIVE!
apiToken: string // primitive - SENSITIVE!
}
}
apiKey: string // primitive - SENSITIVE!
}
```
Output:
```
⚠️ SENSITIVE DATA DETECTED:
- PASSWORD (config.credentials.password)
- TOKEN (config.credentials.apiToken)
- API_KEY (apiKey)
```
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
