### Setup Lynkr and Example Dependencies

Source: https://github.com/fast-editor/lynkr/blob/main/examples/vercel-ai-sdk/README.md

Steps to start the Lynkr server, install example dependencies, and run different chat functionalities.

```bash
# 1. Start Lynkr
cd /path/to/Lynkr
npm start

# 2. Install example dependencies
cd examples/vercel-ai-sdk
npm install

# 3. Run examples
npm run chat     # Basic text generation
npm run stream   # Streaming response
npm run tools    # Tool calling
```

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

### Install Lynkr from Source

Source: https://github.com/fast-editor/lynkr/blob/main/INSTALL.md

Clone the Lynkr repository, install dependencies using npm, copy the example environment file, and start the application.

```bash
git clone https://github.com/Fast-Editor/Lynkr.git
cd Lynkr
npm install
cp .env.example .env
npm start
```

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

### Install Lynkr from Source

Source: https://github.com/fast-editor/lynkr/wiki/Emulating-the-Claude-Code-Backend-for-LLM's-hosted-on-Databricks-&-Azure-Anthropic-(with-MCP,-Git-tools,-and-prompt-caching)

Clone the Lynkr repository, install dependencies, and start the application from source.

```bash
git clone https://github.com/vishalveerareddy123/Lynkr.git
cd Lynkr
npm install
npm start
```

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

### Configure and Start Lynkr from Source

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/installation.md

After cloning and installing dependencies, copy the environment file, edit it with provider credentials, and start the server using npm.

```bash
cp .env.example .env
nano .env
npm start
```

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

### Start Lynkr service

Source: https://github.com/fast-editor/lynkr/blob/main/INSTALL.md

Start the Lynkr service after global installation.

```bash
lynkr start
```

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

### Start Lynkr Server

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/claude-code-cli.md

Start the Lynkr server. If installed from source, use 'npm start'. Wait for the confirmation message indicating the server is listening.

```bash
lynkr start
# Or: npm start (if installed from source)

# Wait for: "Server listening at http://0.0.0.0:8081"
```

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

### Basic Docker Compose Setup

Source: https://github.com/fast-editor/lynkr/blob/main/DEPLOYMENT.md

Basic setup for deploying Lynkr using Docker Compose. Includes copying environment file, starting services, viewing logs, and stopping services.

```bash
# Copy and configure environment
cp .env.example .env
nano .env  # Configure your providers

# Start services
docker compose up -d

# View logs
docker compose logs -f lynkr

# Stop services
docker compose down
```

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

### Environment File Setup

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/providers.md

Recommended for production, use a .env file to manage credentials and configuration. Copy the example file and edit with your specific details.

```env
# Copy example file
cp .env.example .env

# Edit with your credentials
nano .env
```

```env
MODEL_PROVIDER=ollama
DATABRICKS_API_BASE=https://your-workspace.databricks.com
DATABRICKS_API_KEY=dapi1234567890abcdef
PORT=8081
LOG_LEVEL=info

# Tier routing (optional — set all 4 to enable)
TIER_SIMPLE=ollama:llama3.2
TIER_MEDIUM=openrouter:openai/gpt-4o-mini
TIER_COMPLEX=databricks:claude-sonnet
TIER_REASONING=databricks:claude-sonnet
```

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

### Install and Run Lynkr

Source: https://github.com/fast-editor/lynkr/blob/main/skills/lynkr/SKILL.md

Install Lynkr globally, set up Ollama and a model, then start the proxy. Point your AI coding tool to the specified local address.

```bash
npm install -g lynkr
lynkr-setup   # Auto-installs Ollama + pulls a model
lynkr          # Start the proxy
```

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

### Install and Configure Lynkr with Ollama

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/installation.md

Install Lynkr globally after setting up Ollama locally. Configure Ollama environment variables for the model provider and specific model, then start Lynkr. Ollama requires a separate installation and service start.

```bash
# Install Ollama first
brew install ollama  # macOS
# Or download from: https://ollama.ai/download

# Start Ollama service
ollama serve

# Pull a model (in separate terminal)
ollama pull llama3.1:8b

# Install Lynkr
npm install -g lynkr

# Configure
export MODEL_PROVIDER=ollama
export OLLAMA_MODEL=llama3.1:8b

# Start
lynkr start
```

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

### Install Lynkr via Docker

Source: https://github.com/fast-editor/lynkr/blob/main/INSTALL.md

Clone the Lynkr repository, navigate to the directory, and start the Docker containers.

```bash
git clone https://github.com/Fast-Editor/Lynkr.git
cd Lynkr
docker-compose up -d
```

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

### Install and Run Ollama

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/providers.md

Installs Ollama via Homebrew, starts the service, pulls a model, and lists available models.

```bash
# Install Ollama
brew install ollama  # macOS
# Or download from: https://ollama.ai/download

# Start Ollama service
olama serve

# Pull a model
olama pull llama3.1:8b

# Verify model is available
olama list
```

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

### Install and Run Ollama

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/embeddings.md

Install Ollama, start the service, and pull an embedding model. Verify the model is available using `ollama list`.

```bash
# 1. Install Ollama (if not already installed)
brew install ollama  # macOS
# Or download from: https://ollama.ai/download

# 2. Start Ollama service
ollama serve

# 3. Pull embedding model (in separate terminal)
ollama pull nomic-embed-text

# 4. Verify model is available
ollama list
# Should show: nomic-embed-text  ...
```

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

### Clone and Install Dependencies

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/contributing.md

Steps to clone the repository, install dependencies, and set up the environment file for development.

```bash
# Fork the repo on GitHub first, then:
git clone https://github.com/YOUR_USERNAME/Lynkr.git
cd Lynkr

# Install dependencies
npm install

# Copy environment template
cp .env.example .env

# Edit .env with your test credentials
nano .env
```

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

### Docker Deployment Example

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/openclaw-integration.md

Example Docker Compose configuration for deploying Lynkr.

```yaml

```

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

### Install Rust and Graphify

Source: https://github.com/fast-editor/lynkr/blob/main/README.md

Install Rust and then build and install the Graphify tool for optional code intelligence features. Ensure Rust is in your PATH.

```bash
# Install Rust if not already installed
curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh
source $HOME/.cargo/env

# Build and install graphify
git clone https://github.com/safishamsi/graphify
cd graphify
cargo build --release
sudo cp target/release/graphify /usr/local/bin/
```

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

### Docker Compose Setup for Lynkr

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/installation.md

Set up and run Lynkr using Docker Compose, which is recommended for production. This includes cloning, configuring .env, and starting services.

```bash
git clone https://github.com/Fast-Editor/Lynkr.git
cd Lynkr
cp .env.example .env
nano .env
docker-compose up -d
```

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

### Install and Configure Lynkr with Databricks

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/installation.md

Install Lynkr globally, set Databricks environment variables for API access, and start the Lynkr service. Ensure you have your Databricks workspace URL and API key.

```bash
npm install -g lynkr

export MODEL_PROVIDER=databricks
export DATABRICKS_API_BASE=https://your-workspace.cloud.databricks.com
export DATABRICKS_API_KEY=dapi1234567890abcdef

lynkr start
```

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

### Install and Configure Lynkr with OpenRouter

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/installation.md

Install Lynkr globally, set OpenRouter environment variables with your API key and desired model, and start the Lynkr service. OpenRouter offers a wide selection of models and requires an API key from their platform.

```bash
npm install -g lynkr

export MODEL_PROVIDER=openrouter
export OPENROUTER_API_KEY=sk-or-v1-your-key
export OPENROUTER_MODEL=anthropic/claude-3.5-sonnet

lynkr start
```

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

### Lynkr Application Initialization and Navigation

Source: https://github.com/fast-editor/lynkr/blob/main/public/dashboard.html

Handles the initial setup of the application, including setting the default page and starting the refresh countdown. Manages page navigation and updates UI elements accordingly.

```javascript
const App = {
  page: 'overview',
  data: {},
  usageWindow: '7d',
  logFilters: {
    provider: '',
    tier: '',
    error: false
  },
  _refreshTimer: null,
  _countdownTimer: null,
  _countdown: 30,
  _charts: {},
  init() {
    const hash = location.hash.slice(1) || 'overview';
    this.navigate(hash, true);
    this._startCountdown();
  },
  navigate(page, initial=false) {
    this.page = page;
    location.hash = page;
    this._updateTabs();
    document.getElementById('content').innerHTML = loadingState();
    this._destroyCharts();
    this.refresh();
  },
  _updateTabs() {
    document.querySelectorAll('[data-page]').forEach(btn => {
      const active = btn.dataset.page === this.page;
      btn.className = active ? 'ring-tab px-4 py-2 rounded-md text-sm font-medium text-white bg-slate-700' : 'ring-tab px-4 py-2 rounded-md text-sm font-medium text-slate-400 hover:text-white hover:bg-slate-700';
    });
  },
  _startCountdown() {
    this._countdown = 30;
    clearInterval(this._countdownTimer);
    this._countdownTimer = setInterval(() => {
      this._countdown--;
      const el = document.getElementById('refresh-countdown');
      if (el) el.textContent = this._countdown + 's';
      if (this._countdown <= 0) {
        this._countdown = 30;
        this.refresh(true); // silent refresh
      }
    }, 1000);
  },
  async refresh(silent=false) {
    this._countdown = 30;
    if (!silent) document.getElementById('content').innerHTML = loadingState();
    const dot = document.getElementById('live-dot');
    if (dot) {
      dot.className = 'status-dot bg-yellow-500';
    }
    try {
      let data;
      if (this.page === 'overview') data = await this._fetch('/dashboard/api/overview');
      if (this.page === '...
```

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

### MCP Code Mode Workflow Example

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/tools.md

Demonstrates the sequential calls made in MCP Code Mode: first to discover tools, then to get specific tool information, and finally to execute a tool.

```json
// Turn 1: Discover tools
{
  "name": "mcp_list_tools",
  "input": { "server_id": "github" }
}
// → Returns: { total_tools: 47, servers: { github: [{ name: "create_issue", description: "..." }, ...] } }
```

```json
// Turn 2: Get schema for the tool you need
{
  "name": "mcp_tool_info",
  "input": { "server_id": "github", "tool_name": "create_issue" }
}
// → Returns: { inputSchema: { title: { type: "string", ... }, body: { ... } } }
```

```json
// Turn 3: Execute
{
  "name": "mcp_execute",
  "input": {
    "server_id": "github",
    "tool_name": "create_issue",
    "arguments": { "title": "Bug fix", "body": "..." }
  }
}
```

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

### Start llama-server

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/installation.md

Starts the llama-server with a specified model and port.

```bash
./llama-server -m qwen2.5-coder-7b-instruct-q4_k_m.gguf --port 8080
```

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

### Arrange-Act-Assert Pattern Example

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/testing.md

Demonstrates the Arrange-Act-Assert pattern for structuring unit tests in JavaScript, showing setup, execution, and assertion steps.

```javascript
it("should calculate total correctly", () => {
  // Arrange
  const items = [1, 2, 3];

  // Act
  const total = calculateTotal(items);

  // Assert
  assert.strictEqual(total, 6);
});
```

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

### Install Lynkr from Source

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/claude-code-cli.md

Clone the Lynkr repository from GitHub, navigate into the directory, and install dependencies using npm.

```bash
git clone https://github.com/Fast-Editor/Lynkr.git
cd Lynkr && npm install
```

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

### Configure Environment Variables

Source: https://github.com/fast-editor/lynkr/blob/main/DEPLOYMENT.md

Copies the example environment file and opens it for editing.

```bash
sudo cp .env.example .env
sudo nano .env
```

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

### Deploy Lynkr with Docker Compose

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/faq.md

Clone the Lynkr repository, copy the environment example file, configure your credentials in the .env file, and start the services in detached mode using Docker Compose.

```bash
git clone https://github.com/Fast-Editor/Lynkr.git
cd Lynkr
cp .env.example .env
# Edit .env with your credentials
docker-compose up -d
```

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

### Install Node.js on Ubuntu/Debian

Source: https://github.com/fast-editor/lynkr/blob/main/INSTALL.md

Use the NodeSource script to set up Node.js 20.x and then install Node.js on Ubuntu/Debian.

```bash
curl -fsSL https://deb.nodesource.com/setup_20.x | sudo -E bash -
sudo apt-get install -y nodejs
```

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

### Install and Configure Lynkr with AWS Bedrock

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/installation.md

Install Lynkr globally, configure AWS Bedrock environment variables including API key, region, and model ID, then start the Lynkr service. Bedrock requires enabling in your AWS region and requesting model access.

```bash
npm install -g lynkr

export MODEL_PROVIDER=bedrock
export AWS_BEDROCK_API_KEY=ABSK...your-bedrock-api-key
export AWS_BEDROCK_REGION=us-east-1
export AWS_BEDROCK_MODEL_ID=us.anthropic.claude-3-5-sonnet-20241022-v2:0

lynkr start
```

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

### Install Lynkr via One-line Script

Source: https://github.com/fast-editor/lynkr/blob/main/README.md

Install Lynkr using a convenient one-line curl script. This method downloads and executes the installation script.

```bash
curl -fsSL https://raw.githubusercontent.com/Fast-Editor/Lynkr/main/install.sh | bash
```

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

### Install Lynkr via Homebrew

Source: https://github.com/fast-editor/lynkr/blob/main/INSTALL.md

Tap the Lynkr repository and install it using Homebrew on macOS or Linux.

```bash
brew tap fast-editor/lynkr
brew install lynkr
lynkr start
```

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

### Production OpenRouter Setup for Lynkr

Source: https://github.com/fast-editor/lynkr/blob/main/README.md

An example .env configuration for a production environment using OpenRouter, optimized for cost savings with smart tier routing. It includes essential provider and server settings.

```bash
# .env - Production OpenRouter Setup

# ============================================ 
# REQUIRED: Provider Configuration
# ============================================ 
MODEL_PROVIDER=openrouter
OPENROUTER_API_KEY=sk-or-v1-your-key-here
FALLBACK_ENABLED=false

# ============================================ 
# REQUIRED: Server Configuration
# ============================================ 
PORT=8081
HOST=0.0.0.0

# ============================================ 
# TIER ROUTING: Smart Cost Optimization

```

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

### Mixed Local and Cloud Model Provider Setup

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/routing.md

Configure environment variables to define default model providers and specific tiers for different complexity levels. This setup routes simple requests locally and complex ones to cloud providers.

```bash
MODEL_PROVIDER=ollama                        # Startup checks + default provider
TIER_SIMPLE=ollama:llama3.2                  # Score 0-25 → Ollama (free, local)
TIER_MEDIUM=openai:gpt-4o                    # Score 26-50 → OpenAI
TIER_COMPLEX=databricks:claude-sonnet-4-5    # Score 51-75 → Databricks
TIER_REASONING=databricks:claude-opus-4-6    # Score 76-100 → Databricks
```

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

### Lynkr .env File Configuration Example

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/installation.md

Example of a .env file for configuring Lynkr, including core settings, provider-specific credentials, tool execution, and memory system options.

```env
# Core Configuration
MODEL_PROVIDER=databricks
PORT=8081
LOG_LEVEL=info
WORKSPACE_ROOT=/path/to/your/projects

# Databricks Configuration
DATABRICKS_API_BASE=https://your-workspace.cloud.databricks.com
DATABRICKS_API_KEY=dapi1234567890abcdef

# Tool Execution
TOOL_EXECUTION_MODE=server

# Memory System (optional)
MEMORY_ENABLED=true
MEMORY_RETRIEVAL_LIMIT=5
```

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

### Basic Test Structure with Setup and Teardown

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/testing.md

Illustrates a more detailed test structure with setup (beforeEach) and teardown (afterEach) logic, including environment variable saving/restoring and module cache clearing for isolated testing.

```javascript
const assert = require("assert");
const { describe, it, beforeEach, afterEach } = require("node:test");

describe("Module Name", () => {
  let originalEnv;

  beforeEach(() => {
    // Save original environment
    originalEnv = { ...process.env };

    // Clear module cache to get fresh config
    delete require.cache[require.resolve("../src/config")];
  });

  afterEach(() => {
    // Restore original environment
    process.env = originalEnv;
  });

  it("should handle valid input", () => {
    // Arrange
    const input = "test-input";

    // Act
    const result = myFunction(input);

    // Assert
    assert.strictEqual(result, "expected-output");
  });

  it("should throw error for invalid input", () => {
    // Assert that function throws
    assert.throws(
      () => myFunction(null),
      /Expected error message/
    );
  });
});
```

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

### Embeddings Configuration Examples

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/api.md

Examples of environment variables for configuring embeddings providers, including Ollama (local) and OpenAI (cloud).

```bash
# Ollama embeddings (local)
EMBEDDINGS_PROVIDER=ollama
OLLAMA_API_BASE=http://localhost:11434
OLLAMA_EMBEDDINGS_MODEL=nomic-embed-text

# OpenAI embeddings (cloud)
EMBEDDINGS_PROVIDER=openai
OPENAI_API_KEY=sk-...
OPENAI_EMBEDDINGS_MODEL=text-embedding-3-small
```

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

### Warm Up Ollama After Startup

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/troubleshooting.md

Send a test request to a model after Lynkr starts to pre-load it into memory and reduce the chance of a cold start on the first user request.

```bash
curl http://localhost:8081/v1/messages \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-3-5-sonnet","max_tokens":10,"messages":[{"role":"user","content":"hi"}]}'
```

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

### Writing Unit Tests with Node.js Test Runner

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/contributing.md

Example of how to write unit tests using Node.js's built-in test runner, including setup, execution, and assertion.

```javascript
const assert = require("assert");
const { describe, it, beforeEach, afterEach } = require("node:test");

describe("Feature name", () => {
  beforeEach(() => {
    // Setup
  });

  afterEach(() => {
    // Cleanup
  });

  it("should do something specific", () => {
    // Arrange
    const input = "test";

    // Act
    const result = myFunction(input);

    // Assert
    assert.strictEqual(result, "expected");
  });
});
```

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

### Start Ollama Server

Source: https://github.com/fast-editor/lynkr/blob/main/README.md

Start the Ollama server using the 'ollama serve' command. This is necessary to resolve the 'connect ECONNREFUSED' error when using Ollama with Lynkr.

```bash
ollama serve
```

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

### Start Lynkr

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/headroom.md

Run this command to start the Lynkr service, which will automatically manage the Headroom container.

```bash
npm start
```

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

### Install Lynkr Globally

Source: https://github.com/fast-editor/lynkr/wiki/Emulating-the-Claude-Code-Backend-for-LLM's-hosted-on-Databricks-&-Azure-Anthropic-(with-MCP,-Git-tools,-and-prompt-caching)

Install the Lynkr CLI globally using npm for easy access.

```bash
npm install -g lynkr
lynkr start
```

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

### Set Credentials and Start Lynkr

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/troubleshooting.md

If Lynkr fails to start due to missing credentials, ensure the MODEL_PROVIDER and the corresponding API key are set. If the port is in use, find and kill the existing process or specify a different port.

```bash
# Check provider is configured
echo $MODEL_PROVIDER
echo $DATABRICKS_API_KEY  # or other provider key

# If empty, set them:
export MODEL_PROVIDER=databricks
export DATABRICKS_API_KEY=your-key
```

```bash
# Find process using port 8081
lsof -i :8081

# Kill the process
kill -9 <PID>

# Or use different port
export PORT=8082
lynkr start
```

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

### Install and Build llama.cpp

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/embeddings.md

Clone the llama.cpp repository and build it. Optional GPU support can be enabled during the build process.

```bash
# 1. Clone and build llama.cpp
git clone https://github.com/ggerganov/llama.cpp
cd llama.cpp

# Build with GPU support (optional):
# For CUDA (NVIDIA): make LLAMA_CUDA=1
# For Metal (Apple Silicon): make LLAMA_METAL=1
# For CPU only: make
make

# 2. Download embedding model (GGUF format)
# Example: nomic-embed-text GGUF
wget https://huggingface.co/nomic-ai/nomic-embed-text-v1.5-GGUF/resolve/main/nomic-embed-text-v1.5.Q4_K_M.gguf

# 3. Start llama-server with embedding model
./llama-server \
  -m nomic-embed-text-v1.5.Q4_K_M.gguf \
  --port 8080 \
  --embedding

# 4. Verify server is running
curl http://localhost:8080/health
# Should return: {"status":"ok"}
```

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

### Health Check Integration Example

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/headroom.md

Example JSON response from the /health/ready endpoint, showing Headroom status within overall system health.

```json
{
  "status": "ready",
  "checks": {
    "database": { "healthy": true },
    "memory": { "healthy": true },
    "headroom": {
      "healthy": true,
      "enabled": true,
      "service": "available",
      "docker": "running"
    }
  }
}
```

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

### Install MLX OpenAI Server and Dependencies

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/providers.md

Installs the MLX OpenAI Server using pip and optionally installs ffmpeg for audio transcription support via Homebrew.

```bash
# Create virtual environment
python3.11 -m venv .venv
source .venv/bin/activate

# Install
pip install mlx-openai-server

# Optional: for audio transcription
brew install ffmpeg
```

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

### Click-to-Copy Install Command

Source: https://github.com/fast-editor/lynkr/blob/main/docs/index.html

JavaScript code for a click-to-copy functionality for the install command. It listens for clicks on an element with the ID 'install-cmd' and copies 'npm install -g lynkr' to the clipboard, providing visual feedback.

```javascript
// Click-to-copy install command const installEl = document.getElementById('install-cmd'); const copyHint = document.getElementById('copy-hint'); if (installEl) { const copy = () => { navigator.clipboard.writeText('npm install -g lynkr').then(() => { copyHint.textContent = 'copied!'; copyHint.classList.add('copied'); setTimeout(() => { copyHint.textContent = 'click to copy'; copyHint.classList.remove('copied'); }, 2000); }); }; installEl.addEventListener('click', copy); installEl.addEventListener('keydown', e => { if (e.key === 'Enter' || e.key === ' ') copy(); }); }
```

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

### Minimal Ollama Setup for Lynkr

Source: https://github.com/fast-editor/lynkr/blob/main/README.md

A copy-paste ready .env configuration for a minimal working setup using Ollama. This includes required provider, Ollama, server, and compatibility settings, along with recommended performance options.

```bash
# .env - Minimal Ollama Setup

# ============================================ 
# REQUIRED: Provider Configuration
# ============================================ 
MODEL_PROVIDER=ollama
FALLBACK_ENABLED=false

# ============================================ 
# REQUIRED: Ollama Settings
# ============================================ 
OLLAMA_ENDPOINT=http://localhost:11434
OLLAMA_MODEL=qwen2.5-coder:latest

# ============================================ 
# REQUIRED: Server Configuration
# ============================================ 
PORT=8081
HOST=0.0.0.0

# ============================================ 
# REQUIRED: Claude Code/Cursor Compatibility
# ============================================ 
POLICY_MAX_STEPS=50
POLICY_MAX_TOOL_CALLS=100
POLICY_SAFE_COMMANDS_ENABLED=false

# ============================================ 
# OPTIONAL: Performance (Recommended)
# ============================================ 
LOG_LEVEL=warn
LOAD_SHEDDING_ENABLED=true
LOAD_SHEDDING_HEAP_THRESHOLD=0.85
```

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

### Azure OpenAI Full Endpoint URL Example

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/providers.md

Provides an example of the required full endpoint URL format for Azure OpenAI, including resource name, deployment path, and API version.

```text
https://your-resource.openai.azure.com/openai/deployments/gpt-4o/chat/completions?api-version=2025-01-01-preview
```

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

### Multi-line Prompt Example

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/claude-code-cli.md

Demonstrates how to use the Claude CLI for complex, multi-line prompts, such as generating a function with specific requirements.

```bash
claude "Write a function that:
- Takes an array of numbers
- Filters out even numbers
- Returns the sum of odd numbers"
```

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

### Example MCP Tool Usage

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/tools.md

This JSON object shows how to invoke an MCP tool with its input parameters.

```json
{
  "name": "my-custom-tool",
  "input": {
    "input": "test value"
  }
}
```

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

### Install NVIDIA Container Toolkit

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/docker.md

Installs the NVIDIA Container Toolkit on Ubuntu/Debian systems. This is a prerequisite for enabling GPU support in Docker containers.

```bash
# Ubuntu/Debian
distribution=$(. /etc/os-release;echo $ID$VERSION_ID)
curl -s -L https://nvidia.github.io/nvidia-docker/gpgkey | sudo apt-key add -
curl -s -L https://nvidia.github.io/nvidia-docker/$distribution/nvidia-docker.list | \
  sudo tee /etc/apt/sources.list.d/nvidia-docker.list

sudo apt-get update
sudo apt-get install -y nvidia-container-toolkit
sudo systemctl restart docker
```

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

### Hybrid Setup: Ollama for Local Embeddings and Tiered Chat

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/cursor-integration.md

Configure a hybrid setup using Ollama for local, private embeddings and tier-based routing for chat. This balances cost and performance by using local models for simple requests and cloud for complex ones.

```bash
# Chat: Tier-based routing (set all 4 to enable)
TIER_SIMPLE=ollama:llama3.2
TIER_MEDIUM=openrouter:openai/gpt-4o-mini
TIER_COMPLEX=databricks:databricks-claude-sonnet-4-5
TIER_REASONING=databricks:databricks-claude-sonnet-4-5
FALLBACK_ENABLED=true
FALLBACK_PROVIDER=databricks
DATABRICKS_API_BASE=https://your-workspace.databricks.com
DATABRICKS_API_KEY=your-key

# Embeddings: Ollama (local, private)
OLLAMA_EMBEDDINGS_MODEL=nomic-embed-text

# Cost: Mostly FREE (Ollama handles 70-80% of simple requests)
#       Only complex/reasoning requests go to Databricks
```

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

### Start Lynkr Services with Docker Compose

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/docker.md

Start all defined services in the docker-compose.yml file in detached mode. This is the recommended way to run Lynkr.

```bash
docker-compose up -d
```

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

### Start Lynkr Proxy and Connect Claude Code CLI

Source: https://github.com/fast-editor/lynkr/blob/main/docs/index.md

Start the Lynkr proxy server and then configure the Claude Code CLI to use the local proxy. This allows your CLI tools to communicate through Lynkr.

```bash
# Start the proxy server
npm start  # → Running at http://localhost:8081

# Connect Claude Code CLI
export ANTHROPIC_BASE_URL=http://localhost:8081
export ANTHROPIC_API_KEY=dummy
claude
```

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

### Installing Lynkr Dependencies

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/faq.md

Install project dependencies using npm. The --force flag can be used if necessary, but it's generally recommended to resolve dependency conflicts directly.

```bash
npm install
# Or: npm install -g lynkr --force
```

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

### Create File with Claude CLI

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/claude-code-cli.md

Instruct the Claude CLI to create a new file, such as a server setup.

```bash
# Create file
claude "Create a new Express server in server.js"
```

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

### Run Lynkr in Development Mode

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/contributing.md

Commands to start Lynkr in development mode, with options for auto-restart or normal execution.

```bash
# Auto-restart on file changes
npm run dev

# Or normal mode
npm start
```

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

### Build and Run llama.cpp Server

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/providers.md

Clones the llama.cpp repository, builds the server, downloads a GGUF model, and starts the local server.

```bash
# Clone and build llama.cpp
git clone https://github.com/ggerganov/llama.cpp
cd llama.cpp && make

# Download a GGUF model (example: Qwen2.5-Coder-7B-Instruct-GGUF)
wget https://huggingface.co/Qwen/Qwen2.5-Coder-7B-Instruct-GGUF/resolve/main/qwen2.5-coder-7b-instruct-q4_k_m.gguf

# Start llama-server
./llama-server -m qwen2.5-coder-7b-instruct-q4_k_m.gguf --port 8080

# Verify server is running
curl http://localhost:8080/health
```

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

### Start MLX OpenAI Server with a General Purpose Model

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/providers.md

Launches the MLX OpenAI Server using a general-purpose model.

```bash
# General purpose
mlx-openai-server launch --model-path mlx-community/Qwen2.5-3B-Instruct-4bit --model-type lm
```

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

### Test Lynkr with Claude Code

Source: https://github.com/fast-editor/lynkr/blob/main/INSTALL.md

Verify the installation by sending a simple prompt to Lynkr via the Claude CLI.

```bash
export ANTHROPIC_BASE_URL=http://localhost:8081
export ANTHROPIC_API_KEY=dummy
claude "say hello"
```

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

### Configure Provider Credentials

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/installation.md

Troubleshoot server startup issues by ensuring provider credentials are set.

```bash
# Check your provider is configured
echo $MODEL_PROVIDER
echo $DATABRICKS_API_KEY  # or other provider key

# If empty, set them:
export MODEL_PROVIDER=databricks
export DATABRICKS_API_KEY=your-key
```

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

### Visual Setup Summary for Cursor

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/cursor-integration.md

A visual representation of the required settings in Cursor IDE for OpenAI API integration with Lynkr.

```text
┌─────────────────────────────────────────────────────────┐
│         Cursor Settings → Models → OpenAI API           │
├─────────────────────────────────────────────────────────┤
│                                                         │
│  API Key:     sk-lynkr                                 │
│               (or any non-empty value)                  │
│                                                         │
│  Base URL:    http://localhost:8081/v1                │
│               ⚠️ Must include /v1                      │
│                                                         │
│  Model:       claude-3.5-sonnet                        │
│               (or your provider's model)                │
│                                                         │
└─────────────────────────────────────────────────────────┘
```

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

### Basic Chat Example

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/claude-code-cli.md

Send a simple text prompt to Claude Code CLI to get a response. This demonstrates basic conversational functionality.

```bash
# Simple question
claude "Explain async/await in JavaScript"
```

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

### Ollama Configuration for 100% Local Setup

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/cursor-integration.md

Use this configuration for maximum privacy and cost savings, running all models and embeddings locally. Ensure Ollama and the specified models are installed.

```env
MODEL_PROVIDER=ollama
OLLAMA_MODEL=llama3.1:8b
OLLAMA_EMBEDDINGS_MODEL=nomic-embed-text
```

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

### Route53 Latency-Based Routing Configuration

Source: https://github.com/fast-editor/lynkr/blob/main/DEPLOYMENT.md

Configure AWS Route53 for latency-based multi-region deployment. This example shows a basic A record setup pointing to regional ALBs with health checks enabled.

```yaml
# Route53 configuration (AWS)
Type: A
Name: lynkr.example.com
Value:
  - us-east-1-alb.example.com
  - eu-west-1-alb.example.com
Routing Policy: Latency-based
Health Check: Enabled
```

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

### Quick Start with Docker Compose

Source: https://github.com/fast-editor/lynkr/blob/main/DEPLOYMENT.md

The fastest way to deploy Lynkr in production using Docker Compose. Includes cloning, configuration, and verification steps.

```bash
# 1. Clone and configure
git clone https://github.com/Fast-Editor/Lynkr.git
cd Lynkr
cp .env.example .env
# Edit .env with your credentials

# 2. Deploy with Docker Compose (recommended for most users)
docker compose up -d

# 3. Verify health
curl http://localhost:8080/health
```

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

### Verify Lynkr Installation

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/installation.md

Check if the Lynkr installation was successful by running the version command. This should be done after installing via NPM or Homebrew.

```bash
lynkr --version
```

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

### Verify Graphify Installation

Source: https://github.com/fast-editor/lynkr/blob/main/README.md

Verify that the Graphify tool has been installed correctly by checking its version.

```bash
graphify --version
```

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

### Install Node.js Dependencies

Source: https://github.com/fast-editor/lynkr/blob/main/DEPLOYMENT.md

Installs production dependencies for the Lynkr project using npm.

```bash
npm ci --omit=dev
```

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

### Install Node.js on macOS

Source: https://github.com/fast-editor/lynkr/blob/main/INSTALL.md

Use Homebrew to install Node.js version 20 or higher on macOS.

```bash
brew install node
```

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

### Configure Environment Variables

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/docker.md

Copy the example environment file and edit it with your provider credentials. This file is essential for configuring Lynkr's behavior.

```bash
cp .env.example .env
# Edit .env with your provider credentials
```

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

### Complete Kubernetes Deployment Example

Source: https://github.com/fast-editor/lynkr/blob/main/PERFORMANCE-REPORT.md

A comprehensive YAML definition for deploying the Lynkr application, including resource requests, limits, probes, and environment variables.

```yaml
apiVersion: apps/v1
kind: Deployment
metadata:
  name: lynkr
  namespace: lynkr
  labels:
    app: lynkr
    version: v1.0.0
spec:
  replicas: 3
  strategy:
    type: RollingUpdate
    rollingUpdate:
      maxSurge: 1
      maxUnavailable: 0
  selector:
    matchLabels:
      app: lynkr
  template:
    metadata:
      labels:
        app: lynkr
        version: v1.0.0
      annotations:
        prometheus.io/scrape: "true"
        prometheus.io/port: "8080"
        prometheus.io/path: "/metrics/prometheus"
    spec:
      containers:
      - name: lynkr
        image: your-registry.com/lynkr:v1.0.0
        ports:
        - containerPort: 8080
          name: http
          protocol: TCP
        env:
        - name: PORT
          value: "8080"
        - name: MODEL_PROVIDER
          value: "databricks"
        - name: DATABRICKS_API_BASE
          valueFrom:
            secretKeyRef:
              name: lynkr-secrets
              key: DATABRICKS_API_BASE
        - name: DATABRICKS_API_KEY
          valueFrom:
            secretKeyRef:
              name: lynkr-secrets
              key: DATABRICKS_API_KEY
        - name: PROMPT_CACHE_ENABLED
          value: "true"
        - name: METRICS_ENABLED
          value: "true"
        - name: HEALTH_CHECK_ENABLED
          value: "true"
        - name: GRACEFUL_SHUTDOWN_TIMEOUT
          value: "30000"
        - name: LOAD_SHEDDING_HEAP_THRESHOLD
          value: "0.90"
        - name: CIRCUIT_BREAKER_FAILURE_THRESHOLD
          value: "5"
        resources:
          requests:
            cpu: 500m
            memory: 512Mi
          limits:
            cpu: 2000m
            memory: 2Gi
        livenessProbe:
          httpGet:
            path: /health/live
            port: 8080
          initialDelaySeconds: 10
          periodSeconds: 10
          timeoutSeconds: 5
          failureThreshold: 3
        readinessProbe:
          httpGet:
            path: /health/ready
            port: 8080
          initialDelaySeconds: 5
          periodSeconds: 5
          timeoutSeconds: 3
          failureThreshold: 2
        lifecycle:
          preStop:
            exec:
              command:
              - /bin/sh
              - -c
              - sleep 15
      terminationGracePeriodSeconds: 45
---
apiVersion: v1
kind: Service
metadata:
  name: lynkr
  namespace: lynkr
  labels:
    app: lynkr
spec:
  type: ClusterIP
  ports:
  - port: 8080
    targetPort: 8080
    protocol: TCP
    name: http
  selector:
    app: lynkr
---
apiVersion: v1
kind: Service
metadata:
  name: lynkr-metrics
  namespace: lynkr
  labels:
    app: lynkr
spec:
  type: ClusterIP
  ports:
  - port: 8080
    targetPort: 8080
    protocol: TCP
    name: metrics
  selector:
    app: lynkr

```

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

### Install Ollama on macOS/Linux

Source: https://github.com/fast-editor/lynkr/blob/main/INSTALL.md

Install Ollama using the provided curl script for macOS and Linux systems.

```bash
# macOS/Linux
curl -fsSL https://ollama.com/install.sh | sh
```

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

### Configure Lynkr with Environment Variables

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/installation.md

Set up Lynkr by exporting environment variables for your chosen model provider. This is a quick way to configure before starting the server.

```bash
export MODEL_PROVIDER=databricks
export DATABRICKS_API_BASE=https://your-workspace.databricks.com
export DATABRICKS_API_KEY=your-key
lynkr start
```

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

### Start Docker Compose Services

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/docker.md

Command to start all services defined in the docker-compose.yml file in detached mode.

```bash
# Start services
docker-compose up -d
```

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

### Example MCP Server Configuration

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/tools.md

This JSON object defines a custom MCP tool, including its name, description, and input schema.

```json
{
  "name": "my-custom-tool",
  "description": "Does something useful",
  "inputSchema": {
    "type": "object",
    "properties": {
      "input": {
        "type": "string",
        "description": "Input parameter"
      }
    },
    "required": ["input"]
  }
}
```

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

### Start Ollama and Pull Model

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/installation.md

Starts the Ollama service and pulls the specified model, preparing it for local use.

```bash
ollama serve
ollama pull llama3.2
```

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

### Project Context Example

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/memory-system.md

Demonstrates how project-specific facts, like API port numbers, are stored and injected into relevant conversations to provide context.

```text
User: "This API uses port 3000"
System: [Stores: fact, importance 0.75]

Later...
User: "How do I test the API?"
System: [Injects: "API uses port 3000"]
Assistant: "curl http://localhost:3000/endpoint"
```

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

### Code Explanation Example

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/claude-code-cli.md

Provide a code file as input to Claude Code CLI for explanation. This utilizes the file reading capabilities.

```bash
# Code explanation
claude "Explain this function" < app.js
```

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

### Start Lynkr Gateway

Source: https://github.com/fast-editor/lynkr/blob/main/docs/index.html

Start the Lynkr gateway to enable connections from AI tools. The gateway listens on http://localhost:8081 by default.

```bash
# Start the gateway
$ npm start  # → http://localhost:8081
```

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

### Docker Compose High Availability: Load Balancer Example

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/docker.md

Example of configuring an Nginx load balancer that depends on the Lynkr service.

```yaml
services:
  nginx:
    image: nginx:alpine
    ports:
      - "80:80"
    volumes:
      - ./nginx.conf:/etc/nginx/nginx.conf:ro
    depends_on:
      - lynkr
```

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

### MCP Code Mode Example Workflow

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/token-optimization.md

Demonstrates a typical interaction flow using MCP Code Mode, involving listing tools, inspecting a specific tool's schema, and executing it. This workflow replaces sending all tool schemas with a few meta-tools.

```json
Turn 1: mcp_list_tools({ server_id: "github" })
  → Returns: ["create_issue", "list_prs", "merge_pr", ...]

Turn 2: mcp_tool_info({ server_id: "github", tool_name: "create_issue" })
  → Returns: { inputSchema: { title: string, body: string, ... } }

Turn 3: mcp_execute({
    server_id: "github",
    tool_name: "create_issue",
    arguments: { title: "Bug", body: "..." }
  })
```

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

### Node.js SDK Example: Non-streaming and Streaming

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/api.md

Examples using the Anthropic Node.js SDK to make non-streaming and streaming message requests.

```javascript
import Anthropic from "@anthropic-ai/sdk";

const client = new Anthropic({
  apiKey: "dummy",  // Any value works
  baseURL: "http://localhost:8081"
});

// Non-streaming
const message = await client.messages.create({
  model: "claude-3-5-sonnet-20241022",
  max_tokens: 1024,
  messages: [
    { role: "user", content: "What is the capital of France?" }
  ]
});

console.log(message.content[0].text);

// Streaming
const stream = await client.messages.create({
  model: "claude-3-5-sonnet-20241022",
  max_tokens: 1024,
  messages: [
    { role: "user", content: "Tell me a story" }
  ],
  stream: true
});

for await (const event of stream) {
  if (event.type === "content_block_delta") {
    process.stdout.write(event.delta.text);
  }
}
```

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

### Start MLX OpenAI Server with Advanced Options

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/providers.md

Launches the MLX OpenAI Server with specific configurations for host, port, concurrency, and context length to allow remote connections and optimize performance.

```bash
mlx-openai-server launch \
  --model-path mlx-community/Qwen2.5-Coder-7B-Instruct-4bit \
  --model-type lm \
  --host 0.0.0.0 \
  --port 8000 \
  --max-concurrency 2 \
  --context-length 4096
```

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

### Clone Lynkr Repository and Install Dependencies

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/installation.md

Clone the Lynkr repository from GitHub for development purposes. Install project dependencies using npm.

```bash
git clone https://github.com/Fast-Editor/Lynkr.git
cd Lynkr
npm install
```

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

### Build llama.cpp and Download GGUF Model

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/installation.md

Clone the llama.cpp repository, build the project, and download a GGUF model file. This setup is for maximum performance and control with custom models.

```bash
# Install and build llama.cpp
git clone https://github.com/ggerganov/llama.cpp
cd llama.cpp && make

# Download a GGUF model
wget https://huggingface.co/Qwen/Qwen2.5-Coder-7B-Instruct-GGUF/resolve/main/qwen2.5-coder-7b-instruct-q4_k_m.gguf
```

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

### Python SDK Example: Non-streaming and Streaming

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/api.md

Examples using the OpenAI Python SDK to make non-streaming and streaming chat completion requests.

```python
from openai import OpenAI

client = OpenAI(
    api_key="sk-lynkr",
    base_url="http://localhost:8081/v1"
)

# Non-streaming
response = client.chat.completions.create(
    model="claude-3.5-sonnet",
    messages=[
        {"role": "user", "content": "What is the capital of France?"}
    ],
    max_tokens=1024
)

print(response.choices[0].message.content)

# Streaming
stream = client.chat.completions.create(
    model="claude-3.5-sonnet",
    messages=[
        {"role": "user", "content": "Tell me a story"}
    ],
    max_tokens=1024,
    stream=True
)

for chunk in stream:
    if chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="")
```

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

### Navigation and Initialization

Source: https://github.com/fast-editor/lynkr/blob/main/public/dashboard.html

Handles navigation based on URL hash changes and initializes the application. The application navigates to the route specified by the hash or defaults to 'overview'.

```javascript
window.addEventListener('hashchange', () => App.navigate(location.hash.slice(1) || 'overview'));
App.init();
```

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

### Test Edge Cases Example

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/testing.md

Provides examples of testing edge cases in JavaScript, including handling empty arrays, null input, and negative numbers.

```javascript
it("should handle empty array", () => {
  assert.strictEqual(calculateTotal([]), 0);
});

it("should handle null input", () => {
  assert.throws(() => calculateTotal(null));
});

it("should handle negative numbers", () => {
  assert.strictEqual(calculateTotal([-1, -2]), -3);
});
```

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

### Custom Directory Install Lynkr via Script

Source: https://github.com/fast-editor/lynkr/blob/main/documentation/installation.md

Install Lynkr using the curl script to a custom directory. Specify the desired path using the --dir flag.

```bash
curl -fsSL https://raw.githubusercontent.com/Fast-Editor/Lynkr/main/install.sh | bash -s -- --dir /opt/lynkr
```

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

### Configure Tier Routing for Cost Optimization

Source: https://github.com/fast-editor/lynkr/blob/main/README.md

Set up environment variables in a .env file to define different AI model tiers for various task complexities. This allows Lynkr to use cheaper models for simple tasks and more powerful models for complex ones, saving costs.

```bash
# .env file
MODEL_PROVIDER=ollama
FALLBACK_ENABLED=false

# Use small/fast models for simple tasks
TIER_SIMPLE=ollama:qwen2.5:3b

# Use medium models for normal coding
TIER_MEDIUM=ollama:qwen2.5:7b

# Use powerful models for complex architecture
TIER_COMPLEX=ollama:deepseek-r1:14b
TIER_REASONING=ollama:deepseek-r1:14b

# Optional: Limits (remove for unlimited) for long conversations
POLICY_MAX_STEPS=50
POLICY_MAX_TOOL_CALLS=100
```

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

### Local LLM Setup with Ollama

Source: https://github.com/fast-editor/lynkr/blob/main/docs/index.md

Configure Lynkr to use a local LLM provider like Ollama. Ensure Ollama is running and the specified model is available.

```bash
export MODEL_PROVIDER=ollama
export OLLAMA_MODEL=qwen2.5-coder:latest
```