docs(skill): update react-doctor with CLI specs and analysis workflows

skills/react-doctor/SKILL.md

@@ -1,19 +1,77 @@
 ---
 name: react-doctor
 description: Run after making React changes to catch issues early. Use when reviewing code, finishing a feature, or fixing bugs in a React project.
-version: 1.0.0
+metadata: {
+  version: 1.1.0
+}
 ---
 
 # React Doctor
 
 Scans your React codebase for security, performance, correctness, and architecture issues. Outputs a 0-100 score with actionable diagnostics.
 
+## When to Use
+
+- After modifying hook logic (`useEffect`, `useMemo`, custom hooks).
+- When detecting rendering slowdowns or performance regressions.
+- Before committing changes to ensure a score > 75.
+
 ## Usage
 
 ```bash
-npx -y react-doctor@latest . --verbose --diff
+npx -y react-doctor@latest [directory] [options] [command] 
 ```
 
-## Workflow
+### Main Commands
+- **Full Diagnosis (Automatic)**: `npx -y react-doctor@latest . --verbose --yes` (Scans all files and skips prompts)
+- **Quick Diagnosis (Uncommitted only)**: `npx -y react-doctor@latest . --verbose` (Then answer 'Y' to the prompt)
+- **Score Only**: `npx -y react-doctor@latest . --score`
+- **Auto-fix (via Ami)**: `npx -y react-doctor@latest . --fix`
+
+### Analysis Rules
+
+#### 1. Full Analysis (Complete Project)
+To perform a **complete analysis** of the entire codebase:
+- **Option A (Automatic)**: Use the `--yes` or `-y` flag.
+  - Command: `npx -y react-doctor@latest . --verbose --yes`
+- **Option B (Interactive)**: Run without `-y` and when prompted:
+  - Prompt: `? Found X uncommitted changed files. Only scan current changes? › (Y/n)`
+  - Response: **`n`** (No, I want to scan everything).
+
+#### 2. Quick Analysis (Uncommitted Changes)
+To perform a **fast analysis** only on the files you have modified but not yet committed:
+1. Run: `npx -y react-doctor@latest . --verbose` (Omit the `-y` flag)
+2. When prompted: `? Found X uncommitted changed files. Only scan current changes? › (Y/n)`
+3. Respond with: **`Y`** (Yes, only scan current changes).
+*This is the recommended workflow for rapid iteration during development.*
+
+### Arguments
+- `directory`: project directory to scan (default: ".")
+
+### Options
+- `-v, --version`: display the version number
+- `--lint`: enable linting
+- `--no-lint`: skip linting
+- `--dead-code`: enable dead code detection
+- `--no-dead-code`: skip dead code detection
+- `--verbose`: show file details per rule
+- `--score`: output only the score
+- `-y, --yes`: skip prompts and scan the entire codebase (recommended for full audits)
+- `--project <name>`: select workspace project (comma-separated for multiple)
+- `--diff [base]`: scan only files changed vs base branch
+- `--offline`: skip telemetry (anonymous, not stored, only used to calculate score)
+- `--ami`: enable Ami-related prompts
+- `--fail-on <level>`: exit with error code on diagnostics: error, warning, none (default: "none")
+- `--fix`: open Ami to auto-fix all issues
+- `-h, --help`: display help for command
+
+## Workflow (Guidelines)
 
-Run after making changes to catch issues early. Fix errors first, then re-run to verify the score improved.
+1. **Iteration Strategy**: Use **Quick Analysis** while coding to fix immediate issues. Use **Full Analysis** before final commits or PRs.
+2. **Scan & Analyze**: Execute the diagnosis and prioritize Errors over Warnings.
+3. **Threshold**: If the score is < 75, you MUST prioritize fixing:
+   - **Security**: `eval` usage, exposed secrets.
+   - **Correctness**: Missing keys in lists, asynchronous closures.
+4. **Auto-remediation**: Use the "Ami" tool (`--fix`) for automatic corrections when available.
+5. **Architectural Integrity**: Always verify that refactoring doesn't break modular architecture (e.g., avoid components > 300 lines).
+6. **Validation**: Re-run the scan after fixing to verify the score improvement.