### Install XibeCode from Source

Source: https://github.com/iotserver24/xibecode/blob/main/README.md

Clone the repository, install dependencies, build, and link the CLI globally.

```bash
git clone https://github.com/iotserver24/xibecode
cd xibecode
pnpm install
pnpm run build
pnpm link --global --dir packages/cli
```

--------------------------------

### GitHub Actions CI Setup for XibeCode

Source: https://github.com/iotserver24/xibecode/blob/main/DOCS.md

Integrate XibeCode into a GitHub Actions workflow. This example installs XibeCode, configures it using secrets, and runs a task.

```yaml
- name: Run XibeCode task
  env:
    ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}
    GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
  run: |
    npm install -g xibecode
    xibecode config --set-key "$ANTHROPIC_API_KEY"
    xibecode config --set-url "https://api.anthropic.com"
    xibecode config --set-model "claude-3-7-sonnet-20250219"
    xibecode run "Fix lint errors across the codebase"
```

--------------------------------

### Install MCP Server

Source: https://github.com/iotserver24/xibecode/blob/main/DOCS.md

Install an MCP server to provide the agent with access to external tools. Use `xibecode mcp install` followed by the package name.

```bash
xibecode mcp install @smithery/mcp-postgres
```

--------------------------------

### Minimal Sandbox Setup for XibeCode

Source: https://github.com/iotserver24/xibecode/blob/main/DOCS.md

Install XibeCode globally and configure it non-interactively for use in sandbox environments. This sets the API key, URL, and default model.

```bash
npm install -g xibecode

# Set config (non-interactive)
xibecode config --set-key "$ANTHROPIC_API_KEY"
xibecode config --set-url "https://api.anthropic.com"
xibecode config --set-model "claude-3-7-sonnet-20250219"

# Run a task
xibecode run "your task"
```

--------------------------------

### Install XibeCode CLI

Source: https://github.com/iotserver24/xibecode/blob/main/docs/README.md

Installs the XibeCode command-line interface globally. Use this for end-user installation.

```bash
pnpm install -g xibecode
```

--------------------------------

### Install XibeCode CLI

Source: https://github.com/iotserver24/xibecode/blob/main/DOCS.md

Installs the XibeCode command-line interface globally using npm or pnpm. Verify the installation by checking the version.

```bash
npm install -g xibecode
```

```bash
pnpm add -g xibecode
```

```bash
xibecode -v   # → 0.6.1
```

--------------------------------

### Start Xibecode UI Server

Source: https://github.com/iotserver24/xibecode/blob/main/packages/cli/README.md

Run this command in your terminal to start the Xibecode web UI server.

```bash
xibecode ui
```

--------------------------------

### Install and Build XibeCode Packages

Source: https://github.com/iotserver24/xibecode/blob/main/README.md

Commands to install dependencies, build all packages using Turborepo, and enable watch mode for development.

```bash
pnpm install
pnpm run build          # Build all packages with Turborepo
pnpm run dev            # Watch mode for development
pnpm run sync-version   # Sync versions across all packages
```

--------------------------------

### Verify Installation and Run Tests

Source: https://github.com/iotserver24/xibecode/blob/main/CONTRIBUTING.md

Test the command-line interface (CLI) and run the project's test suite to ensure the installation and build were successful.

```bash
npm run dev -- --help
```

```bash
npm test
```

--------------------------------

### Run Task with Postgres MCP Server

Source: https://github.com/iotserver24/xibecode/blob/main/DOCS.md

After installing and configuring an MCP server (e.g., Postgres), you can run tasks that utilize its tools. This example lists tables and their schemas.

```bash
xibecode run "List all tables in the database and show their schemas"
```

--------------------------------

### Start XibeCode WebUI

Source: https://github.com/iotserver24/xibecode/blob/main/packages/cli/README.md

Start the XibeCode WebUI in your browser. Use the --open flag to automatically open it, or specify a custom port with -p.

```bash
xibecode ui              # Start on localhost:3847
xibecode ui --open       # Auto-open browser
xibecode ui -p 8080      # Custom port
```

--------------------------------

### Install XibeCode from Source

Source: https://github.com/iotserver24/xibecode/blob/main/packages/cli/README.md

Clone the XibeCode repository, install dependencies, build the project, and link it locally. This is useful for development or using the latest unreleased changes.

```bash
git clone https://github.com/iotserver24/xibecode
cd xibecode
pnpm install
pnpm run build
npm link
```

--------------------------------

### Install Project Dependencies

Source: https://github.com/iotserver24/xibecode/blob/main/docs/README.md

Installs all project dependencies using pnpm. Required for contributors after cloning.

```bash
pnpm install
```

--------------------------------

### MCP Server Configuration Example

Source: https://github.com/iotserver24/xibecode/blob/main/DOCS.md

An example JSON structure for configuring an MCP server, specifying the command, arguments, and environment variables.

```json
{
  "my-db": {
    "command": "npx",
    "args": ["-y", "@smithery/mcp-postgres"],
    "env": { "DATABASE_URL": "postgres://..." }
  }
}
```

--------------------------------

### Workflow: Add Authentication

Source: https://github.com/iotserver24/xibecode/blob/main/FEATURES.md

Example of a complex development task broken down into sequential AI-assisted steps, including package installation, middleware creation, and route protection.

```bash
xibecode run "Add JWT authentication:
1. Install jsonwebtoken and bcrypt
2. Create auth middleware
3. Add /login and /register endpoints
4. Protect existing routes
5. Add tests"
```

```text
Iteration 1: run_command("npm install jsonwebtoken bcrypt")
Iteration 2: write_file("middleware/auth.js") # Creates middleware
Iteration 3: read_file("app.js") # Understand current routes
Iteration 4: edit_file("app.js") # Add auth routes
Iteration 5: edit_file("app.js") # Protect routes
Iteration 6: write_file("test/auth.test.js") # Add tests
Iteration 7: run_command("npm test") # Verify
```

--------------------------------

### Install XibeCode via npm

Source: https://github.com/iotserver24/xibecode/blob/main/README.md

Install XibeCode globally using npm or pnpm.

```bash
pnpm install -g xibecode
# or
npm install -g xibecode
```

--------------------------------

### Manage Configuration Settings

Source: https://github.com/iotserver24/xibecode/blob/main/README.md

Examples of using `xibecode config` to set API keys, models, and view current settings.

```bash
xibecode config --set-key YOUR_KEY
xibecode config --set-model claude-sonnet-4-5-20250929
xibecode config --show
```

--------------------------------

### Vitest Test Organization Example

Source: https://github.com/iotserver24/xibecode/blob/main/CODING_STANDARDS.md

Organize tests using describe blocks for grouping and beforeEach for setup. Use 'it' blocks for individual test cases, covering happy paths, edge cases, and error conditions.

```typescript
import { describe, it, expect, beforeEach } from 'vitest';

describe('FunctionName', () => {
  // Setup
  beforeEach(() => {
    // Reset state
  });

  // Happy path
  it('should return expected result for valid input', () => {
    const result = functionName('valid');
    expect(result).toBe('expected');
  });

  // Edge cases
  it('should handle empty input', () => {
    expect(() => functionName('')).toThrow();
  });

  it('should handle null input', () => {
    expect(() => functionName(null)).toThrow();
  });

  // Error cases
  it('should throw error for invalid input', () => {
    expect(() => functionName('invalid')).toThrow(ValidationError);
  });
});
```

--------------------------------

### Search and Install MCP Servers from Smithery

Source: https://context7.com/iotserver24/xibecode/llms.txt

Find and install MCP server packages from the Smithery registry. Use keywords to search.

```bash
xibecode mcp search postgres
```

```bash
xibecode mcp install @modelcontextprotocol/server-postgres
```

--------------------------------

### Build Project

Source: https://github.com/iotserver24/xibecode/blob/main/docs/README.md

Builds the entire XibeCode monorepo using Turborepo. Run this after installing dependencies as a contributor.

```bash
pnpm run build
```

--------------------------------

### Run Autonomous Workflow Examples

Source: https://github.com/iotserver24/xibecode/blob/main/README.md

Examples of using the `xibecode run` command for various tasks, including verbose output and file input.

```bash
xibecode run "Build a REST API with Express"
xibecode run "Fix the TypeScript errors" --verbose
xibecode run --file task.txt
```

--------------------------------

### Install Project Dependencies

Source: https://github.com/iotserver24/xibecode/blob/main/CONTRIBUTING.md

Install project dependencies using either npm or pnpm. This step is crucial before building or running the project.

```bash
npm install
```

```bash
pnpm install
```

--------------------------------

### SkillManager: Load, Search, and Install Skills

Source: https://context7.com/iotserver24/xibecode/llms.txt

Handles loading markdown skills from local directories and remote registries. Use `loadSkills` to initialize, `buildDefaultSkillsPromptForTask` to get relevant skills for a prompt, `installFromSkillsSh` to add new skills, and `learnFromDocs` to synthesize skills from URLs.

```typescript
import { SkillManager } from 'xibecode-core';

const sm = new SkillManager(
  process.cwd(),
  process.env.ANTHROPIC_API_KEY,
  undefined,         // custom base URL (optional)
  'claude-sonnet-4-5-20250929',
  'anthropic'
);
await sm.loadSkills();

// List all loaded skills
const all = sm.listSkills();
console.log(all.map(s => `${s.name} [${s.provenance}]`));

// Get task-specific bundled prompt (keyword-matched subset)
const promptBlock = await sm.buildDefaultSkillsPromptForTask(
  'Write Playwright E2E tests for the checkout flow',
  process.cwd()
);
// Use promptBlock in AgentConfig.defaultSkillsPrompt

// Search locally loaded skills
const matches = sm.searchSkills('docker');

// Install a skill from skills.sh
const installed = await sm.installFromSkillsSh('playwright-e2e');
console.log(installed.success, installed.filePath);

// Learn a skill from any documentation URL (AI-synthesized)
const result = await sm.learnFromDocs(
  'prisma',
  'https://www.prisma.io/docs',
  25,
  (msg) => console.log(msg)
);
console.log(`Scraped ${result.pagesScraped} pages → ${result.filePath}`);

// Custom skill file format (.xibecode/skills/my-skill.md):
/*
---
name: my-skill
description: Custom coding pattern for this project
tags: architecture, api
---

## Rules
- Always validate inputs with Zod
- Use Result<T, E> instead of throwing
*/

```

--------------------------------

### Auto-discover and Install Skills

Source: https://github.com/iotserver24/xibecode/blob/main/DOCS.md

The agent can automatically discover and install skills based on the task. It searches for relevant skills, installs them, and then uses their instructions.

```bash
xibecode run "Set up a Next.js project with best practices"
```

--------------------------------

### Start XibeCode Chat Interface

Source: https://github.com/iotserver24/xibecode/blob/main/docs/README.md

Launches the interactive chat interface for XibeCode.

```bash
xibecode chat
```

--------------------------------

### Install XibeCode CLI

Source: https://github.com/iotserver24/xibecode/blob/main/packages/cli/README.md

Install the XibeCode CLI globally using npm. This command makes the xibecode executable available in your terminal.

```bash
npm install -g xibecode
```

--------------------------------

### Start WebUI and Pentest Automation

Source: https://github.com/iotserver24/xibecode/blob/main/implementation_plan.md

Use these commands to start the XibeCode WebUI and the agent-browser for penetration testing automation. Ensure the WebUI is running on the specified port before launching the automation script.

```bash
# Terminal 1: Start WebUI
xibecode ui --port 3847

# Terminal 2: Run pentest automation
agent-browser open http://localhost:3847
```

--------------------------------

### Event-Driven Architecture Example

Source: https://github.com/iotserver24/xibecode/blob/main/ARCHITECTURE.md

Example of using agent events for UI updates or error handling. Listen for 'tool_call' or 'error' events.

```typescript
agent.on('tool_call', (data) => { /* Update UI */ });
agent.on('error', (error) => { /* Handle error */ });
```

--------------------------------

### Configure Environment Variables

Source: https://github.com/iotserver24/xibecode/blob/main/CONTRIBUTING.md

Copy the example environment file and add your Anthropic API key. This is necessary for certain functionalities, especially related to AI features.

```bash
cp .env.example .env

echo "ANTHROPIC_API_KEY=your_api_key_here" >> .env
```

--------------------------------

### Install TypeDoc Dependencies

Source: https://github.com/iotserver24/xibecode/blob/main/docs/README.md

Installs TypeDoc and its Markdown plugin as development dependencies. Required for generating API documentation.

```bash
pnpm add -D typedoc typedoc-plugin-markdown
```

--------------------------------

### Plugin Pattern Example

Source: https://github.com/iotserver24/xibecode/blob/main/ARCHITECTURE.md

Dynamic loading of plugins and registration of their tools. Assumes `loadPlugins` and `registerTools` functions exist.

```typescript
const plugins = await loadPlugins();
plugins.forEach(p => registerTools(p.tools));
```

--------------------------------

### Changelog Format Example

Source: https://github.com/iotserver24/xibecode/blob/main/CONTRIBUTING.md

Example format for the CHANGELOG.md file, detailing changes for a specific version including added, changed, fixed, deprecated, and removed items.

```markdown
## [0.3.6] - 2026-02-14

### Added
- New feature X
- Tool Y support

### Changed
- Improved performance of Z

### Fixed
- Bug in component A
- Issue with B

### Deprecated
- Old API method C

### Removed
- Deprecated feature D
```

--------------------------------

### Run Autonomous Workflow with PR Creation Examples

Source: https://github.com/iotserver24/xibecode/blob/main/README.md

Examples of using `xibecode run-pr` for autonomous coding with automatic branch and PR creation, including specifying a branch and draft status.

```bash
xibecode run-pr "Fix the TypeScript errors"
xibecode run-pr "Add input validation" --branch feat/validation --draft
```

--------------------------------

### Install and Use Browser MCP Server

Source: https://github.com/iotserver24/xibecode/blob/main/DOCS.md

Install and use an MCP server for browser automation, such as Playwright. This allows the agent to interact with web pages.

```bash
xibecode mcp install @smithery/mcp-playwright
```

```bash
xibecode run "Take a screenshot of https://example.com and describe the layout"
```

--------------------------------

### JSON Configuration File Example

Source: https://github.com/iotserver24/xibecode/blob/main/README.md

Configure API keys, models, and other settings via a JSON file located at ~/.xibecode/config.json. This provides an alternative to environment variables.

```json
{
  "apiKey": "sk-ant-...",
  "model": "claude-sonnet-4-5-20250929",
  "provider": "anthropic",
  "maxIterations": 50,
  "costMode": "normal"
}
```

--------------------------------

### Build and Develop Project

Source: https://github.com/iotserver24/xibecode/blob/main/CONTRIBUTING.md

Build all packages using Turborepo or run the project in watch mode for continuous development. Ensure dependencies are installed first.

```bash
pnpm install

pnpm run build

pnpm run dev
```

--------------------------------

### Loop Detection Example (With Detection)

Source: https://github.com/iotserver24/xibecode/blob/main/FEATURES.md

Shows how XibeCode tracks actions and detects repetitive calls to prevent infinite loops.

```javascript
// XibeCode tracks recent actions:
History: [
  { tool: "read_file", input: "app.js", time: 1000 },
  { tool: "read_file", input: "app.js", time: 1100 },
  { tool: "read_file", input: "app.js", time: 1200 },
]

// After 3 identical calls:
⚠️  Loop detected: read_file called 3+ times with same parameters
```

--------------------------------

### Configure XibeCode API Key

Source: https://github.com/iotserver24/xibecode/blob/main/docs/README.md

Sets your API key for XibeCode. This is a required step after installation for users.

```bash
xibecode config --set-key YOUR_API_KEY
```

--------------------------------

### Façade Pattern Example

Source: https://github.com/iotserver24/xibecode/blob/main/ARCHITECTURE.md

Simplified public API for agent interaction. Initializes an `EnhancedAgent` and runs an agent task.

```typescript
const agent = new EnhancedAgent(config);
await agent.runAgent('Build a login page');
```

--------------------------------

### Internal Implementation Detail Example

Source: https://github.com/iotserver24/xibecode/blob/main/CODING_STANDARDS.md

Implementation details within private members do not require JSDoc documentation.

```typescript
// OK: Implementation internals don't need JSDoc
private _internalCache = new Map();
```

--------------------------------

### Define a Custom Plugin with a Tool

Source: https://github.com/iotserver24/xibecode/blob/main/API_REFERENCE.md

Example of a JavaScript plugin defining a custom tool 'analyze_performance' that takes a file path as input.

```javascript
// my-custom-plugin.js
export default {
  name: 'custom-plugin',
  version: '1.0.0',
  tools: [
    {
      name: 'analyze_performance',
      description: 'Analyze code performance',
      input_schema: {
        type: 'object',
        properties: {
          file: { type: 'string' }
        }
      },
      execute: async ({ file }) => {
        // Analysis logic
        return { success: true, output: 'Performance analysis results' };
      }
    }
  ]
};
```

--------------------------------

### Workflow: Refactor to TypeScript

Source: https://github.com/iotserver24/xibecode/blob/main/FEATURES.md

Demonstrates the AI-driven process for converting a JavaScript project to TypeScript, including dependency installation, configuration, and file conversion.

```bash
xibecode run "Convert src/ to TypeScript"
```

```text
Iteration 1: search_files("src/**/*.js") # Find all JS files
Iteration 2: read_file("package.json") # Check dependencies
Iteration 3: run_command("npm install -D typescript @types/node")
Iteration 4: write_file("tsconfig.json") # Create config
Iteration 5-15: For each file:
  - read_file("file.js")
  - write_file("file.ts") # Convert with types
Iteration 16: run_command("npx tsc") # Verify compilation
```

--------------------------------

### Loop Detection Example (Without Detection)

Source: https://github.com/iotserver24/xibecode/blob/main/FEATURES.md

Illustrates a scenario that would lead to an infinite loop without XibeCode's loop detection mechanism.

```javascript
// Without loop detection:
Iteration 1: read_file("app.js")
Iteration 2: read_file("app.js")  // Same thing again
Iteration 3: read_file("app.js")  // Still doing it
Iteration 4: read_file("app.js")  // Infinite loop!
```

--------------------------------

### Run Task with OpenRouter

Source: https://github.com/iotserver24/xibecode/blob/main/DOCS.md

Execute a task using any model available through OpenRouter. This example uses Gemini 2.5 Pro. Requires the OPENROUTER_API_KEY environment variable.

```bash
xibecode run \
  --provider openai \
  --api-key "$OPENROUTER_API_KEY" \
  --base-url "https://openrouter.ai/api/v1" \
  --model "google/gemini-2.5-pro-exp-03-25" \
  "your task"
```

--------------------------------

### Tip: Be Specific About Files

Source: https://github.com/iotserver24/xibecode/blob/main/FEATURES.md

Illustrates the importance of specifying exact file paths for clarity and accuracy when instructing XibeCode.

```bash
# Good
xibecode run "Add error handling to src/api/users.js"

# Less good (AI might guess wrong file)
xibecode run "Add error handling to the users file"
```

--------------------------------

### Teach Project Conventions with xibecode

Source: https://github.com/iotserver24/xibecode/blob/main/DOCS.md

Use this command to instruct the agent on project conventions, such as preferred libraries and file structures. Ensure the agent understands these guidelines for consistent development.

```bash
xibecode run "Remember: we use Zod for validation, Prisma for DB, pnpm for packages, vitest for tests, and all API routes live in src/routes/"
```

--------------------------------

### Conventional Commit Example: Fix

Source: https://github.com/iotserver24/xibecode/blob/main/CONTRIBUTING.md

Example of a bug fix commit message, detailing the issue and the solution implemented.

```bash
fix(agent): prevent infinite loop in tool execution

Loop detector was not correctly tracking tool calls with
complex input objects. Fixed by using JSON serialization
for comparison.

Fixes #456
```

--------------------------------

### JSDoc Documentation Example

Source: https://github.com/iotserver24/xibecode/blob/main/CONTRIBUTING.md

Document public functions, classes, and interfaces using JSDoc comments, including descriptions, examples, parameters, and exceptions.

```typescript
/**
 * Executes a tool with the given input
 *
 * Validates tool permissions, performs safety checks, and executes
 * the tool in the current mode context.
 *
 * @example
 * ```typescript
 * const result = await executeTool('read_file', { path: '/app.ts' });
 * console.log(result.output);
 * ```
 *
 * @param toolName - Name of the tool to execute
 * @param input - Tool input parameters
 * @returns Tool execution result
 * @throws {PermissionError} If tool not allowed in current mode
 * @throws {SafetyError} If operation deemed unsafe
 *
 * @see {@link CodingToolExecutor} for tool definitions
 * @since 0.3.0
 * @category Tool Execution
 */
async function executeTool(
  toolName: string,
  input: Record<string, any>
): Promise<ToolResult> {
  // Implementation
}
```

--------------------------------

### Conventional Commit Example: Feature

Source: https://github.com/iotserver24/xibecode/blob/main/CONTRIBUTING.md

Example of a feature commit message, including a detailed body explaining the changes and a footer referencing an issue.

```bash
feat(tools): add http_request tool with retry logic

Add a new http_request tool that supports:
- GET, POST, PUT, DELETE methods
- Custom headers
- Automatic retry on failure
- Timeout configuration

Closes #123
```

--------------------------------

### Build a Single XibeCode Package

Source: https://github.com/iotserver24/xibecode/blob/main/README.md

Commands to navigate into a specific package directory and build it individually.

```bash
cd packages/core && pnpm run build
cd packages/cli && pnpm run build
```

--------------------------------

### Git Commit Message Type Examples

Source: https://github.com/iotserver24/xibecode/blob/main/CODING_STANDARDS.md

Examples of Git commit messages demonstrating the 'feat' and 'fix' types with scope, subject, body, and footer.

```bash
feat(tools): add http_request tool with retry logic

Add a new http_request tool that supports:
- GET, POST, PUT, DELETE methods
- Custom headers
- Automatic retry on failure
- Timeout configuration

Closes #123
```

```bash
fix(agent): prevent infinite loop in tool execution

Loop detector was not correctly tracking tool calls with
complex input objects. Fixed by using JSON serialization
for comparison.

Fixes #456
```

--------------------------------

### Show Full Instructions for a XibeCode Skill

Source: https://context7.com/iotserver24/xibecode/llms.txt

Display the complete instructions and usage details for a specific skill.

```bash
xibecode skills show playwright-e2e
```

--------------------------------

### Information Disclosure and Configuration Tampering via GET /api/config

Source: https://github.com/iotserver24/xibecode/blob/main/pentest-report.md

Exposes sensitive information like API keys and allows unauthenticated modification of configuration via GET requests.

```APIDOC
## GET /api/config

### Description
Retrieves the current application configuration, potentially exposing sensitive information like API keys and base URLs. This endpoint also allows for unauthenticated configuration modification.

### Method
GET

### Endpoint
/api/config

### Response
#### Success Response (200)
- **apiKey** (string) - The current API key.
- **model** (string) - The name of the AI model in use.
- **baseUrl** (string) - The base URL for external API calls.

#### Response Example
```json
{
  "apiKey": "sk-sensitive-key-12345",
  "model": "gpt-3.5-turbo",
  "baseUrl": "https://api.openai.com/v1"
}
```
```

--------------------------------

### Load Complex Tasks from File

Source: https://github.com/iotserver24/xibecode/blob/main/DOCS.md

For intricate tasks, define them in a text file and use the `--file` option with `xibecode run`. The `--verbose` flag provides detailed output.

```bash
cat > task.txt << 'EOF'
Refactor the authentication module:
1. Extract token validation into a separate utility
2. Add refresh token support
3. Write unit tests for all auth functions
4. Update the README with new auth flow docs
EOF
xibecode run --file task.txt --verbose
```

--------------------------------

### Use Multiple Config Profiles

Source: https://context7.com/iotserver24/xibecode/llms.txt

Manage different configuration profiles for distinct environments or projects. Use the --profile flag to switch between them.

```bash
xibecode config --profile work --set-key sk-...
```

```bash
xibecode run "Fix the auth bug" --profile work
```

--------------------------------

### Clone Repository and Set Up Remotes

Source: https://github.com/iotserver24/xibecode/blob/main/CONTRIBUTING.md

Clone your forked repository and add the upstream remote to fetch changes from the main project. Ensure you replace YOUR_USERNAME with your GitHub username.

```bash
git clone https://github.com/YOUR_USERNAME/xibecode.git
cd xibecode

git remote add upstream https://github.com/iotserver24/xibecode.git
```

--------------------------------

### Initialize and Use MCPClient

Source: https://github.com/iotserver24/xibecode/blob/main/API_REFERENCE.md

Instantiate MCPClient and connect to an MCP server using a specified type and configuration. List available tools and execute a remote tool with parameters.

```typescript
import { MCPClient } from 'xibecode';

const mcp = new MCPClient();

// Connect to MCP server
await mcp.connect('filesystem', {
  type: 'stdio',
  command: 'mcp-server-filesystem',
  args: ['/workspace']
});

// List available tools
const tools = await mcp.listTools('filesystem');

// Execute remote tool
const result = await mcp.executeTool('filesystem', 'read_file', {
  path: '/data.json'
});
```

--------------------------------

### Example Test Suite Output

Source: https://github.com/iotserver24/xibecode/blob/main/packages/cli/README.md

An example of a generated test suite using Vitest for a TypeScript function. It includes tests for successful execution, type checking, handling empty arrays, and error conditions.

```typescript
import { describe, it, expect, vi } from 'vitest';
import { calculateTotal } from '../utils/helpers';

describe('calculateTotal', () => {
  it('should execute calculateTotal successfully', () => {
    expect(calculateTotal([])).toBeDefined();
  });

  it('should return correct type from calculateTotal', () => {
    expect(typeof calculateTotal([])).toBe('number');
  });

  it('should handle empty array', () => {
    expect(calculateTotal([])).toBe(0);
  });

  it('should handle errors in calculateTotal', () => {
    expect(() => calculateTotal(undefined)).toThrow();
  });
});
```

--------------------------------

### Start Interactive Terminal Chat

Source: https://github.com/iotserver24/xibecode/blob/main/packages/cli/README.md

Launches an interactive terminal chat session for tool use. This command also automatically starts the WebUI, accessible at http://localhost:3847, enabling real-time synchronization between the TUI and WebUI.

```bash
# Start with both TUI and WebUI
xibecode chat

# WebUI opens automatically at http://localhost:3847
```

--------------------------------

### Basic Agent Initialization and Run

Source: https://github.com/iotserver24/xibecode/blob/main/API_REFERENCE.md

Initializes an EnhancedAgent with API key and model, then runs a task to create a React component.

```typescript
import { EnhancedAgent } from 'xibecode';

const agent = new EnhancedAgent({
  apiKey: process.env.ANTHROPIC_API_KEY!,
  model: 'claude-sonnet-4-5-20250929',
  mode: 'agent'
});

await agent.runAgent('Create a new React component for user profile');
```

--------------------------------

### Simple Getter Example

Source: https://github.com/iotserver24/xibecode/blob/main/CODING_STANDARDS.md

Getters and setters that are obvious in their functionality do not require JSDoc documentation.

```typescript
// OK: No JSDoc for simple accessors
get name(): string {
  return this._name;
}
```

--------------------------------

### Private Helper Function Example

Source: https://github.com/iotserver24/xibecode/blob/main/CODING_STANDARDS.md

Private helper functions that are self-explanatory do not require JSDoc documentation.

```typescript
// OK: No JSDoc needed for obvious helpers
private isValid(str: string): boolean {
  return str.length > 0;
}
```

--------------------------------

### Tip: Break Down Huge Tasks

Source: https://github.com/iotserver24/xibecode/blob/main/FEATURES.md

Advises on decomposing large, complex development goals into smaller, manageable steps for more effective AI assistance.

```bash
# Instead of:
xibecode run "Build complete e-commerce platform"

# Do:
xibecode run "Create product model and CRUD endpoints"
xibecode run "Add shopping cart functionality"
xibecode run "Implement checkout process"
```

--------------------------------

### Naming Conventions: Classes, Interfaces, and Types

Source: https://github.com/iotserver24/xibecode/blob/main/CODING_STANDARDS.md

Use PascalCase for classes, interfaces, and types. Examples include descriptive nouns for classes and specific configurations or modes for interfaces and types.

```typescript
// ✅ Good: Descriptive nouns
class EnhancedAgent { }
class FileEditor { }
class ContextManager { }

// ❌ Bad: Vague or abbreviated
class Agt { }
class Mgr { }
class Util { }
```

--------------------------------

### Manage MCP Servers

Source: https://github.com/iotserver24/xibecode/blob/main/README.md

Commands for managing MCP servers, including listing, adding, searching, and installing skills.

```bash
xibecode mcp list
xibecode mcp add my-server --command "node" --args "server.js"
xibecode mcp search                # Search Smithery registry
xibecode mcp install <skill-id>    # Install from Smithery
```

--------------------------------

### Start Interactive Chat with Specific Model

Source: https://github.com/iotserver24/xibecode/blob/main/README.md

Initiate an interactive chat session specifying a particular AI model.

```bash
xibecode chat --model claude-opus-4-5-20251101
```

--------------------------------

### Cross-Platform Command Execution (Windows)

Source: https://github.com/iotserver24/xibecode/blob/main/FEATURES.md

Demonstrates how XibeCode executes commands on Windows, using `dir` for file listing.

```bash
xibecode run "List all Python files"

# AI knows it's Windows and uses:
run_command({ command: "dir /s /b *.py" })
```

--------------------------------

### Strategy Pattern Example

Source: https://github.com/iotserver24/xibecode/blob/main/ARCHITECTURE.md

Runtime selection of an editing strategy based on edit type. Requires a `selectEditStrategy` function.

```typescript
const strategy = selectEditStrategy(editType);
await strategy.apply(file, changes);
```

--------------------------------

### Integration Tests with Vitest

Source: https://github.com/iotserver24/xibecode/blob/main/CONTRIBUTING.md

Example of integration tests for the EnhancedAgent using Vitest. Verifies the complete workflow execution of the agent.

```typescript
import { describe, it, expect } from 'vitest';
import { EnhancedAgent } from 'xibecode-core';

describe('EnhancedAgent Integration', () => {
  it('should execute complete workflow', async () => {
    const agent = new EnhancedAgent(config);
    const result = await agent.runAgent('task');
    expect(result.success).toBe(true);
  });
});
```

--------------------------------

### Show Current XibeCode Configuration

Source: https://github.com/iotserver24/xibecode/blob/main/DOCS.md

Displays the current configuration settings for XibeCode, including API keys, URLs, and models.

```bash
xibecode config --show
```

--------------------------------

### Sensitive Data Exposure via GET /api/files/read

Source: https://github.com/iotserver24/xibecode/blob/main/pentest-report.md

Allows reading any file within the working directory without authentication.

```APIDOC
## GET /api/files/read

### Description
Reads the content of a specified file within the working directory without requiring authentication.

### Method
GET

### Endpoint
/api/files/read

### Query Parameters
- **path** (string) - Required - The path to the file to read.

### Response
#### Success Response (200)
- **content** (string) - The content of the file.

#### Response Example
```json
{
  "content": "This is the content of the file."
}
```
```

--------------------------------

### Run XibeCode CLI Command

Source: https://github.com/iotserver24/xibecode/blob/main/docs/README.md

Executes a task using the XibeCode CLI. Replace 'Build a REST API' with your desired task.

```bash
xibecode run "Build a REST API"
```