evizi-kit 1.0.0

package/kits/shared/docs/ai-code-review.md

@@ -0,0 +1,440 @@
+# AI Code Review Skills
+
+Automated code review skills for AI agents to perform PR reviews and self-reviews before creating pull requests.
+
+## 📚 Learn About Agent Skills
+
+Before using these skills, we recommend learning about how agent skills work:
+
+👉 **[Agent Skills with Anthropic - DeepLearning.AI](https://learn.deeplearning.ai/courses/agent-skills-with-anthropic/lesson/ldn5c3/introduction)**
+
+This free course covers:
+- How skills work and best practices for creating them
+- Progressive disclosure: Skills are loaded into context only when needed
+- Using skills with different agents (Claude AI, Claude Code, Claude Agent SDK)
+- Combining skills with MCP and sub-agents
+
+## 🎯 Available Skills
+
+This repository contains two code review skills: [AI Code Review](https://gitea.evizi.com/ai/design/src/branch/ai-code-review/)
+
+### 1. PR Review Skill (`pr-review`)
+Reviews pull request code changes and creates inline comments on Git platforms (GitHub, Gitea, GitLab, Bitbucket).
+
+**Triggers on:**
+- "Review PR #42"
+- "Please review pull request 123"
+- "Can you check PR number 15?"
+- "Review the changes in pull request #8"
+
+📖 **[Full Documentation](./documents/pr-review-agent/document.md)**
+
+### 2. Self Review Skill (`self-review`)
+Performs automated pre-PR code review with a clear Ready/Not Ready verdict for PR creation.
+
+**Triggers on:**
+- "Self-review for my code changes"
+- "Check my code before PR"
+- "Review my changes"
+- "Is my code ready for PR?"
+- "self-review"
+
+📖 **[Full Documentation](./documents/self-review-agent/document.md)**
+
+## 🚀 Quick Start
+
+### Installation
+
+**Option 1: Clone these skills to your project**
+
+**For Cursor IDE:**
+
+*macOS/Linux:*
+```bash
+# Navigate to your project directory
+
+# Clone and install skills
+rm -rf temp-skills
+git clone -b ai-code-review https://gitea.evizi.com/ai/design.git temp-skills
+mkdir -p .cursor
+cp -r temp-skills/skills .cursor/
+cp temp-skills/project.config.json .cursor/
+rm -rf temp-skills
+```
+
+*Windows (PowerShell):*
+```powershell
+# Navigate to your project directory
+
+# Clone and install skills
+Remove-Item -Recurse -Force temp-skills -ErrorAction SilentlyContinue
+git clone -b ai-code-review https://gitea.evizi.com/ai/design.git temp-skills
+New-Item -ItemType Directory -Force -Path .cursor | Out-Null
+Copy-Item -Recurse -Force temp-skills\skills .cursor\
+Copy-Item -Force temp-skills\project.config.json .cursor\
+Remove-Item -Recurse -Force temp-skills
+```
+
+**For GitHub Copilot:**
+
+*macOS/Linux:*
+```bash
+# Navigate to your project directory
+cd /path/to/your/project
+
+# Clone and install skills
+rm -rf temp-skills
+git clone -b ai-code-review https://gitea.evizi.com/ai/design.git temp-skills
+mkdir -p .github
+cp -r temp-skills/skills .github/
+cp -r temp-skills/project.config.json .github/
+rm -rf temp-skills
+```
+
+*Windows (PowerShell):*
+```powershell
+# Navigate to your project directory
+cd C:\path\to\your\project
+
+# Clone and install skills
+Remove-Item -Recurse -Force temp-skills -ErrorAction SilentlyContinue
+git clone -b ai-code-review https://gitea.evizi.com/ai/design.git temp-skills
+New-Item -ItemType Directory -Force -Path .github | Out-Null
+Copy-Item -Recurse -Force temp-skills\skills .github\
+Copy-Item -Force temp-skills\project.config.json .github\
+Remove-Item -Recurse -Force temp-skills
+```
+
+**Option 2: Manual setup**
+
+1. Copy the `skills` folder to your project:
+   - For Cursor IDE: Place in `.cursor/skills/`
+   - For GitHub Copilot: Place in `.github/skills/`
+   - For other agents: Check agent documentation for skill directory location
+
+2. Copy `project.config.json` to the same directory (`.cursor/` or `.github/`)
+
+3. Update the configuration (see Configuration section below)
+
+**Note:** If these folders/files already exist in your project, they will be replaced with the new versions.
+
+### Configuration
+
+1. **Update Configuration File**: Edit `.cursor/project.config.json` (or `.github/project.config.json`):
+
+```json
+{
+  "baseBranch": "main",
+  "platform": "github",
+  "platformConfig": {
+    "github": {
+      "apiUrl": "https://api.github.com",
+      "owner": "your-org",
+      "repo": "your-repo"
+    }
+  },
+  "enableCommentPosting": true,
+  "enableNotification": true
+}
+```
+
+**Configuration Options:**
+- `enableCommentPosting` (boolean, default: `true`) - Enable/disable posting inline comments to the git platform
+- `enableNotification` (boolean, default: `false`) - Enable/disable Google Chat notifications
+
+**Operation Modes:**
+- **Full Mode**: `enableCommentPosting: true` + `enableNotification: true` → Posts comments + sends notification
+- **Comment-Only Mode**: `enableCommentPosting: true` + `enableNotification: false` → Posts comments only
+- **Notification-Only Mode**: `enableCommentPosting: false` + `enableNotification: true` → Sends notification only (no token needed)
+
+2. **Environment Variables**: Set up tokens for your Git platform:
+
+```bash
+# GitHub
+export GITHUB_TOKEN="your-github-token"
+
+# GitLab
+export GITLAB_TOKEN="your-gitlab-token"
+
+# Gitea
+export GITEA_TOKEN="your-gitea-token"
+
+# Bitbucket
+export BITBUCKET_TOKEN="your-bitbucket-token"
+export BITBUCKET_USERNAME="your-bitbucket-username"
+
+# Google Chat notifications (optional)
+export GOOGLE_CHAT_WEBHOOK_URL="your-webhook-url"
+```
+
+3. **Verify Installation**:
+
+```bash
+# Check if skills are installed
+ls .cursor/skills/  # Should show: pr-review, self-review
+
+# Check if config exists
+cat .cursor/project.config.json
+```
+
+4. **Start Using**: Your AI agent will now automatically recognize and use these skills!
+
+## 🤖 Using Skills with Different AI Agents
+
+### Cursor
+
+Cursor natively supports agent skills through its skill system.
+
+**Setup:**
+1. Skills are automatically discovered from `.cursor/skills/` directory
+2. Cursor's AI agent will automatically invoke skills based on your prompts
+
+**Usage Example:**
+```
+You: Review PR #42
+[Cursor AI invokes pr-review skill automatically]
+
+You: Self-review for my code changes
+[Cursor AI invokes self-review skill automatically]
+```
+
+**Skill Location:**
+- `.cursor/skills/pr-review/SKILL.md`
+- `.cursor/skills/self-review/SKILL.md`
+
+### GitHub Copilot
+
+GitHub Copilot Workspace supports agent skills through `.github/skills/` directory.
+
+**Setup:**
+1. Place skills in `.github/skills/` directory
+2. Create `.github/agents/` with agent definitions if needed
+3. GitHub Copilot will recognize and use skills
+
+**Usage Example:**
+```
+You: Review PR #42
+[GitHub Copilot invokes pr-review skill]
+
+You: Check my code before PR
+[GitHub Copilot invokes self-review skill]
+```
+
+**Skill Location:**
+- `.github/skills/pr-review/SKILL.md`
+- `.github/skills/self-review/SKILL.md`
+
+### Claude Code
+
+Claude Code (by Anthropic) supports agent skills natively.
+
+**Setup:**
+1. Skills can be placed in `.cursor/skills/` or custom directory
+2. Configure skill path in Claude Code settings
+3. Claude Code automatically loads skills on startup
+
+**Usage Example:**
+```
+You: Review PR #42
+[Claude Code invokes pr-review skill]
+
+You: Self-review my changes
+[Claude Code invokes self-review skill]
+```
+
+**How it Works:**
+- Claude Code progressively discloses skills
+- Skill descriptions are always in context
+- Full instructions loaded only when needed
+- Combines with MCP tools and sub-agents
+
+### Antigravity
+
+Antigravity supports custom agent skills and workflows.
+
+**Setup:**
+1. Configure skill directory in Antigravity settings
+2. Skills are loaded based on directory structure
+3. Antigravity's orchestrator handles skill invocation
+
+**Usage Example:**
+```
+You: Review pull request 42
+[Antigravity invokes pr-review skill]
+
+You: Is my code ready for PR?
+[Antigravity invokes self-review skill]
+```
+
+**Configuration:**
+- Skills directory: Configurable in Antigravity settings
+- Supports both `.cursor/skills/` and `.github/skills/` patterns
+
+## 📖 Detailed Documentation
+
+### PR Review Skill
+
+The PR review skill performs comprehensive code analysis against domain-specific checklists:
+
+- **Frontend**: 100+ checks (React, Vue, Angular, performance, accessibility)
+- **Backend**: 70+ checks (API design, database, security, OWASP)
+- **Mobile**: 110+ checks (iOS, Android, threading, memory)
+- **Automation**: 40+ checks (E2E tests, selectors, stability)
+
+**Workflow:**
+1. Fetch PR code using Git commands
+2. Detect tech stack and load appropriate checklist
+3. Analyze code changes
+4. Generate standard review output (JSON)
+5. Show preview and ask for confirmation
+6. Post inline comments to Git platform
+7. Send Google Chat notification (optional)
+8. Return to previous branch (cleanup)
+
+**Platform Support:**
+- GitHub
+- GitLab
+- Gitea
+- Bitbucket
+
+[Read Full PR Review Documentation →](./documents/pr-review-agent/document.md)
+
+### Self Review Skill
+
+The self-review skill helps you verify code quality before creating a PR:
+
+**Workflow:**
+1. Compare current branch with target branch
+2. Detect tech stack and load checklist
+3. Analyze changes
+4. Generate table-based report with verdict
+5. Save to `issues.md`
+
+**Verdict Logic:**
+- ✅ **Ready to Create PR** - Zero critical issues
+- ⚠️ **Ready with Warnings** - Zero critical, some warnings/suggestions
+- ❌ **Not Ready for PR** - One or more critical issues
+
+[Read Full Self Review Documentation →](./documents/self-review-agent/document.md)
+
+## 🛠 Advanced Configuration
+
+### Platform-Specific Configuration
+
+**GitHub:**
+```json
+{
+  "platform": "github",
+  "platformConfig": {
+    "github": {
+      "apiUrl": "https://api.github.com",
+      "owner": "your-org",
+      "repo": "your-repo"
+    }
+  },
+  "enableCommentPosting": true,
+  "enableNotification": false
+}
+```
+
+**GitLab:**
+```json
+{
+  "platform": "gitlab",
+  "platformConfig": {
+    "gitlab": {
+      "apiUrl": "https://gitlab.com/api/v4",
+      "projectId": "your-project-id"
+    }
+  },
+  "enableCommentPosting": true,
+  "enableNotification": false
+}
+```
+
+**Gitea:**
+```json
+{
+  "platform": "gitea",
+  "platformConfig": {
+    "gitea": {
+      "apiUrl": "https://gitea.evizi.com/api/v1",
+      "owner": "your-org",
+      "repo": "your-repo"
+    }
+  },
+  "enableCommentPosting": true,
+  "enableNotification": false
+}
+```
+
+**Bitbucket:**
+```json
+{
+  "platform": "bitbucket",
+  "platformConfig": {
+    "bitbucket": {
+      "apiUrl": "https://api.bitbucket.org/2.0",
+      "workspace": "your-workspace",
+      "repoSlug": "your-repo"
+    }
+  },
+  "enableCommentPosting": true,
+  "enableNotification": false
+}
+```
+
+## 📁 Repository Structure
+
+```
+.
+├── .cursor/
+│   ├── skills/
+│   │   ├── pr-review/
+│   │   │   ├── SKILL.md                    # PR review skill definition
+│   │   │   ├── scripts/
+│   │   │   │   └── post-review.py          # Unified dispatcher script
+│   │   │   └── references/
+│   │   │       ├── frontend.md             # Frontend checklist
+│   │   │       ├── backend.md              # Backend checklist
+│   │   │       ├── mobile.md               # Mobile checklist
+│   │   │       └── automation.md           # Automation checklist
+│   │   └── self-review/
+│   │       ├── SKILL.md                    # Self review skill definition
+│   │       ├── templates/
+│   │       │   └── issues.template.md      # Report template
+│   │       └── references/
+│   │           ├── frontend.md
+│   │           ├── backend.md
+│   │           ├── mobile.md
+│   │           └── automation.md
+│   └── project.config.json                 # Project configuration
+├── .github/
+└── └── (Alternative location for GitHub Copilot)
+```
+
+## 🎓 Learning Resources
+
+1. **[Agent Skills with Anthropic - DeepLearning.AI](https://learn.deeplearning.ai/courses/agent-skills-with-anthropic/lesson/ldn5c3/introduction)**
+   - Free course on building and using agent skills
+   - Learn about progressive disclosure and skill architecture
+   - Combine skills with MCP and sub-agents
+
+2. **[PR Review Skill Documentation](./documents/pr-review-agent/document.md)**
+   - Complete guide to PR review workflow
+   - Platform-specific configuration
+   - Architecture diagrams
+
+3. **[Self Review Skill Documentation](./documents/self-review-agent/document.md)**
+   - Pre-PR verification workflow
+   - Verdict logic and severity levels
+   - Example report output
+
+## 🤝 Contributing
+
+Skills are now an open standard with a standardized format that works with any skills-compatible agent. Feel free to:
+
+- Extend existing skills with new checklists
+- Create new skills for different workflows
+- Improve platform support
+- Add new features to the dispatcher script

package/kits/shared/docs/increase-unit-test-coverage.md

@@ -0,0 +1,77 @@
+# Increase Unit Test Coverage Skill
+
+## 1. Objective
+
+The **Increase Unit Test Coverage** skill is designed to help developers and AI agents systematically improve a codebase's test coverage. It solves the problem of overwhelming coverage reports by filtering and prioritizing files that are below a defined threshold, providing a clear, actionable roadmap for improvement.
+
+## 2. Components
+
+- **SKILL.md**: The core instruction set that guides the AI agent through the systematic process of increasing coverage.
+- **filter_low_coverage.py**: A powerful Python script that:
+    - Automatically detects project configuration.
+    - Parses `coverage-summary.json`.
+    - Generates a concise, AI-friendly Markdown report (`low-coverage-report.md`).
+- **project.config.json**: A central configuration file (located in `.cursor/`, `.github/`, `.agent/`, or root) that defines threshold, test commands, directories, and exclusion patterns.
+- **Coverage Report**: The raw output from the project's testing framework (e.g., Jest's `coverage-summary.json`).
+
+## 3. Setup Guide
+
+1.  **Skill Installation**: Ensure the skill folder is located in your IDE's respective directory (e.g., `/Users/dungvo/Workspace/ai-evizi/design/.cursor/skills/increase-unit-test-coverage`).
+2.  **Configuration**: **[CRITICAL]** You MUST update your `project.config.json` with your project-specific settings. The skill relies on these values to run the correct tests and find the right files:
+    ```json
+    {
+      "unitTestCoverage": {
+        "enabled": true,
+        "coverageReportPath": "coverage/coverage-summary.json",
+        "outputReportPath": "low-coverage-report.md",
+        "threshold": 90.0,
+        "reportTopN": 30,
+        "reportShowFullList": false,
+        "testCommand": "npm test -- --coverage",
+        "testDirectory": "src/__tests__",
+        "testFilePattern": "**/*.test.{ts,tsx,js,jsx}",
+        "excludePatterns": ["*.css.ts", "constants.ts"],
+        "excludeDirectories": ["models", "types"],
+        "testingFramework": "jest",
+        "testingLibrary": "react-testing-library"
+      }
+    }
+    ```
+
+### Configuration Reference
+
+| Field | Description |
+| :--- | :--- |
+| `enabled` | Set to `true` to enable the coverage improvement skill. |
+| `coverageReportPath` | Path to the `coverage-summary.json` file generated by your test runner. |
+| `outputReportPath` | (Optional) Path where the filtered `low-coverage-report.md` will be saved. |
+| `threshold` | The target coverage percentage (0.0 to 100.0). Files below this are prioritized. |
+| `reportTopN` | (Optional) Limits the number of files shown in the summary report (default: 30). |
+| `reportShowFullList` | (Optional) If `true`, shows all files in the report regardless of TopN limit. |
+| `testCommand` | The exact command to run tests with coverage reporting (e.g., `npm test -- --coverage`). |
+| `testDirectory` | The directory where your test files are located. |
+| `testFilePattern` | (Optional) Glob pattern for test files (e.g., `**/*.test.ts`). |
+| `excludePatterns` | (Optional) List of file patterns to ignore (e.g., `["*.css.ts", "*.types.ts"]`). |
+| `excludeDirectories`| (Optional) List of directories to ignore (e.g., `["models", "types"]`). |
+| `testingFramework` | The framework used (e.g., `jest`, `vitest`, `mocha`). |
+| `testingLibrary` | The library used (e.g., `react-testing-library`, `enzyme`). |
+3.  **Dependencies**: Ensure **Python 3** is available in your environment to run the filter script.
+
+## 4. How to Use
+
+This skill follows a collaborative workflow where the **User** initiates the task and the **AI Agent** executes the systematic improvement process.
+
+### User Role
+1.  **Initiate Task**: Ask the AI Agent to improve coverage (e.g., *"Increase test coverage for files below 90%"*).
+2.  **Monitor Progress**: Review the AI's actions as it moves through the prioritized list.
+3.  **Confirm Results**: Review the final report once the target threshold is reached.
+
+### AI Agent Role
+1.  **Configuration Check**: Reads `project.config.json` to identify test commands, directories, and thresholds.
+2.  **Generate Baseline**: Runs the project's coverage command (e.g., `npm test -- --coverage`) to get current data.
+3.  **Filtered Reporting**: Executes the `filter_low_coverage.py` script to generate a prioritized `low-coverage-report.md`.
+4.  **Sequential Improvement**:
+    -   **Analyze**: Reads source files and existing tests for prioritized files.
+    -   **Generate Tests**: Writes new test cases for uncovered paths (Zero, Critical, etc.).
+    -   **Verify**: Re-runs coverage after each file to confirm improvement.
+5.  **Iteration**: Continues the cycle until the target coverage threshold is met.