Claude Context

https://github.com/zilliztech/claude-context
Admin
Claude Context is an MCP plugin that adds semantic code search to Claude Code and other AI coding...
Tokens:28,212
Snippets:439
Trust Score:9.8
License:MIT
Update:1 month ago
Show doc for...
Tokens:
Raw
### Windows Setup and Development (PowerShell)

Source: https://github.com/zilliztech/claude-context/blob/master/README.md

Clone the repository, configure git line endings, install dependencies, build, and start development mode on Windows.

```powershell
git clone https://github.com/zilliztech/claude-context.git
cd claude-context
git config core.autocrlf false
pnpm install
pnpm build
pnpm dev
```

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

### Run Example (Bash)

Source: https://github.com/zilliztech/claude-context/blob/master/README.md

Navigate to the basic usage example directory and run it in development mode.

```bash
cd examples/basic-usage
pnpm dev
```

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

### Install Dependencies

Source: https://github.com/zilliztech/claude-context/blob/master/examples/basic-usage/README.md

Install project dependencies using pnpm. This command should be run before executing the example.

```bash
pnpm install
```

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

### Package and Install Extension

Source: https://github.com/zilliztech/claude-context/blob/master/packages/vscode-extension/CONTRIBUTING.md

Steps to package the VSCode extension into a .vsix file and install it locally for testing.

```bash
# Navigate to extension directory
cd packages/vscode-extension

# Package the extension (remove existing .vsix file if present)
pnpm run package

# Uninstall any existing version
code --uninstall-extension semanticcodesearch-xxx.vsix

# Install the packaged extension
code --install-extension semanticcodesearch-xxx.vsix
```

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

### Clone and Install Dependencies (Bash)

Source: https://github.com/zilliztech/claude-context/blob/master/README.md

Clone the repository, navigate into the directory, and install project dependencies using pnpm.

```bash
git clone https://github.com/zilliztech/claude-context.git
cd claude-context
pnpm install
```

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

### Run Claude Context Example

Source: https://github.com/zilliztech/claude-context/blob/master/examples/basic-usage/README.md

Execute the main example script using pnpm. Ensure environment variables are set beforehand.

```bash
pnpm run start
```

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

### Install @zilliz/claude-context-core

Source: https://github.com/zilliztech/claude-context/blob/master/packages/core/README.md

Use npm to install the core package for Claude Context.

```bash
npm install @zilliz/claude-context-core
```

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

### Install Dependencies for Chrome Extension

Source: https://github.com/zilliztech/claude-context/blob/master/packages/chrome-extension/CONTRIBUTING.md

Navigate to the extension's directory and install its dependencies using pnpm.

```bash
cd packages/chrome-extension
pnpm install
```

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

### MCP Server Quick Commands

Source: https://github.com/zilliztech/claude-context/blob/master/packages/mcp/CONTRIBUTING.md

Use these commands for building, developing, and starting the MCP server.

```bash
pnpm build:mcp
```

```bash
pnpm dev:mcp
```

```bash
pnpm start
```

```bash
pnpm start:with-env
```

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

### Start Development Mode

Source: https://github.com/zilliztech/claude-context/blob/master/CONTRIBUTING.md

Start the project in development mode, which typically includes watching for file changes and hot-reloading.

```bash
pnpm dev
```

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

### Build and Run Development Mode (Bash)

Source: https://github.com/zilliztech/claude-context/blob/master/README.md

Build all project packages and start the development server with file watching enabled.

```bash
pnpm build
pnpm dev
```

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

### Start MCP Server Locally

Source: https://github.com/zilliztech/claude-context/blob/master/AGENTS.md

Command to start the MCP server locally using tsx.

```bash
pnpm --filter @zilliz/claude-context-mcp start        # tsx src/index.ts
```

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

### Example Output

Source: https://github.com/zilliztech/claude-context/blob/master/examples/basic-usage/README.md

This is the expected output when running the Claude Context basic usage example. It shows connection status, indexing progress, and search results.

```text
🚀 Claude Context Real Usage Example
===============================
...
🔌 Connecting to vector database at: ...

📖 Starting to index codebase...
🗑️  Existing index found, clearing it first...
📊 Indexing stats: 45 files, 234 code chunks

🔍 Performing semantic search...

🔎 Search: "vector database operations"
   1. Similarity: 89.23%
      File: /path/to/packages/core/src/vectordb/milvus-vectordb.ts
      Language: typescript
      Lines: 147-177
      Preview: async search(collectionName: string, queryVector: number[], options?: SearchOptions)...

🎉 Example completed successfully!
```

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

### Build Core Package with pnpm

Source: https://github.com/zilliztech/claude-context/blob/master/packages/core/CONTRIBUTING.md

Use this command to build the core package. Ensure you have pnpm installed and the project dependencies are set up.

```bash
pnpm build:core
```

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

### Install Dependencies with uv

Source: https://github.com/zilliztech/claude-context/blob/master/evaluation/README.md

Installs project dependencies using uv for Python environments. Ensure you are in the 'evaluation' directory and have activated your virtual environment.

```bash
cd evaluation && uv sync
source .venv/bin/activate
```

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

### Install Node.js Dependencies and Set Environment Variables

Source: https://github.com/zilliztech/claude-context/blob/master/python/README.md

Before running the bridge, ensure Node.js dependencies are installed and necessary environment variables like OPENAI_API_KEY and MILVUS_ADDRESS are set.

```bash
cd .. && pnpm install

export OPENAI_API_KEY="your-openai-api-key"

export MILVUS_ADDRESS="localhost:19530"
```

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

### Install pnpm Globally

Source: https://github.com/zilliztech/claude-context/blob/master/CONTRIBUTING.md

Install the pnpm package manager globally. Ensure it is available in your system's PATH.

```powershell
# Ensure pnpm is installed
npm install -g pnpm
```

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

### Configure Local Milvus Connection

Source: https://github.com/zilliztech/claude-context/blob/master/examples/basic-usage/README.md

Configure Milvus connection details for a local setup. Use the server address and port for `uri`. If authentication is enabled, set the `token` with your username and password.

```bash
export MILVUS_ADDRESS="http://localhost:19530"
export MILVUS_TOKEN="<your_username>:<your_password>"
```

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

### Project Build and Development Commands

Source: https://github.com/zilliztech/claude-context/blob/master/AGENTS.md

Standard pnpm commands for managing the monorepo, including installation, building packages, and development watch modes.

```bash
pnpm install
pnpm build                 # build all packages (examples built last)
pnpm build:core            # build a single package: also build:mcp, build:vscode
pnpm dev                   # watch all; or dev:core / dev:mcp / dev:vscode
pnpm lint                  # eslint across packages; lint:fix to autofix
pnpm typecheck             # tsc --noEmit across packages
pnpm clean                 # rimraf dist in every package
```

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

### Run Chrome Extension in Development Mode

Source: https://github.com/zilliztech/claude-context/blob/master/packages/chrome-extension/CONTRIBUTING.md

Start the extension in watch mode for automatic rebuilds during development.

```bash
pnpm dev  # Watch mode for automatic rebuilds
```

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

### Index Codebase with Ignore Patterns

Source: https://github.com/zilliztech/claude-context/blob/master/packages/mcp/CONTRIBUTING.md

Example of using the `index_codebase` tool with custom ignore patterns. This allows excluding specific files or directories during indexing.

```json
{"path": "/repo/path", "ignorePatterns": ["static/**", "*.tmp"]}
```

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

### Conventional Commit Examples

Source: https://github.com/zilliztech/claude-context/blob/master/CONTRIBUTING.md

Examples of commit messages following the conventional commit format. This format helps in automating changelog generation and versioning.

```git
type(scope): description

feat(core): add new embedding provider
fix(vscode): resolve search result display issue
docs(readme): update installation instructions
refactor(mcp): improve error handling
```

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

### Build Chrome Extension

Source: https://github.com/zilliztech/claude-context/blob/master/packages/chrome-extension/README.md

Build the Chrome extension using pnpm. This command is used during development for manual installation.

```bash
cd packages/chrome-extension
pnpm build
```

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

### Set OpenAI API Key

Source: https://github.com/zilliztech/claude-context/blob/master/examples/basic-usage/README.md

Set your OpenAI API key for embeddings. This is a prerequisite for the example.

```bash
export OPENAI_API_KEY="your-openai-api-key"
```

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

### Configure Background Sync Polling

Source: https://github.com/zilliztech/claude-context/blob/master/packages/mcp/README.md

Controls the startup and periodic background synchronization behavior. Set to false to disable polling and rely on the trigger watcher for multi-instance setups.

```bash
# Default: true. Set to false to disable startup + periodic polling.
CLAUDE_CONTEXT_BACKGROUND_SYNC=false
```

```bash
# Optional: control how often sync runs (default: 300000 = 5 minutes)
CLAUDE_CONTEXT_SYNC_INTERVAL_MS=60000
```

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

### Build and Test Extension in Different Browsers

Source: https://github.com/zilliztech/claude-context/blob/master/packages/chrome-extension/CONTRIBUTING.md

Use this command to build the extension and then load it into various browsers for compatibility testing.

```bash
# Build and test in different browsers
pnpm build
# Load extension in Chrome, Edge, Brave, etc.
```

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

### Build All Packages (Workflow)

Source: https://github.com/zilliztech/claude-context/blob/master/CONTRIBUTING.md

Commands for building all packages in the project. Includes options for a clean build and development mode.

```bash
# Build all packages
pnpm build

# Clean and rebuild
pnpm clean && pnpm build

# Development mode (watch all packages)
pnpm dev
```

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

### Build and Development Commands

Source: https://github.com/zilliztech/claude-context/blob/master/packages/vscode-extension/CONTRIBUTING.md

Commands for building, watching, and packaging the VSCode extension.

```bash
# Build VSCode extension
pnpm build:vscode

# Watch mode for development
pnpm dev:vscode

# Package extension
pnpm package
```

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

### Clone the Repository

Source: https://github.com/zilliztech/claude-context/blob/master/CONTRIBUTING.md

Clone the project repository and navigate into the project directory.

```bash
git clone https://github.com/your-username/claude-context.git
cd claude-context
```

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

### Clean and Build Chrome Extension

Source: https://github.com/zilliztech/claude-context/blob/master/packages/chrome-extension/CONTRIBUTING.md

Run these commands to clean previous builds and create a production-ready bundle for the Chrome extension.

```bash
pnpm clean
pnpm build
```

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

### Grep Method: Initial Directory Listing

Source: https://github.com/zilliztech/claude-context/blob/master/evaluation/case_study/pydata_xarray_6938/README.md

This log illustrates the initial, broad directory listing performed by the grep method to navigate the Xarray repository. It highlights the inefficiency of this approach compared to targeted search.

```log
🔧 list_directory('repo__pydata__xarray')
⚙️ Result: Retrieved directory structure with multiple subdirectories
   Problem: Basic directory navigation without targeted search
```

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

### Running the MCP Server

Source: https://github.com/zilliztech/claude-context/blob/master/packages/mcp/CONTRIBUTING.md

Commands to build and run the MCP server for direct use or with MCP clients.

```bash
pnpm build
```

```bash
pnpm start
```

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

### Configuration Methods

Source: https://github.com/zilliztech/claude-context/blob/master/packages/core/README.md

Methods for configuring the context and its behavior.

```APIDOC
## hasIndex

### Description
Check if codebase is already indexed.

### Signature
`hasIndex(path)

### Parameters
- **path** (string) - Required - The path to the codebase.
```

```APIDOC
## updateIgnorePatterns

### Description
Update ignore patterns.

### Signature
`updateIgnorePatterns(patterns)

### Parameters
- **patterns** (string[]) - Required - An array of glob patterns to ignore.
```

```APIDOC
## addCustomIgnorePatterns

### Description
Add custom ignore patterns.

### Signature
`addCustomIgnorePatterns(patterns)

### Parameters
- **patterns** (string[]) - Required - An array of glob patterns to add to ignore list.
```

```APIDOC
## addCustomExtensions

### Description
Add custom file extensions.

### Signature
`addCustomExtensions(extensions)

### Parameters
- **extensions** (string[]) - Required - An array of file extensions to support.
```

```APIDOC
## updateEmbedding

### Description
Switch embedding provider.

### Signature
`updateEmbedding(embedding)

### Parameters
- **embedding** (object) - Required - The new embedding provider instance.
```

```APIDOC
## updateVectorDatabase

### Description
Switch vector database.

### Signature
`updateVectorDatabase(vectorDB)

### Parameters
- **vectorDB** (object) - Required - The new vector database instance.
```

```APIDOC
## updateSplitter

### Description
Switch code splitter.

### Signature
`updateSplitter(splitter)

### Parameters
- **splitter** (object) - Required - The new code splitter instance.
```

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

### Trigger Re-index with Claude Code Hook

Source: https://github.com/zilliztech/claude-context/blob/master/packages/mcp/README.md

An example of a Claude Code hook configuration that triggers an immediate re-index by touching the sync trigger file after an Edit or Write operation.

```json
"hooks": {
  "PostToolUse": [
    {
      "matcher": "Edit|Write",
      "hooks": [
        {
          "type": "command",
          "command": "touch ~/.context/.sync-trigger"
        }
      ]
    }
  ]
}
```

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

### Enable Debug Mode for MCP Clients

Source: https://github.com/zilliztech/claude-context/blob/master/docs/troubleshooting/troubleshooting-guide.md

Start Claude Code or Gemini CLI with the --debug flag to enable detailed logging. This is useful for diagnosing issues when standard checks fail.

```bash
claude --debug
```

```bash
gemini --debug
```

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

### Verify Bundle Size (Unix/macOS)

Source: https://github.com/zilliztech/claude-context/blob/master/packages/chrome-extension/CONTRIBUTING.md

Check the size of the built distribution files on Unix-like systems.

```bash
ls -lh dist/
```

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

### Xarray swap_dims() Mutation Bug Example

Source: https://github.com/zilliztech/claude-context/blob/master/evaluation/case_study/pydata_xarray_6938/README.md

This Python code demonstrates a bug in Xarray's `swap_dims()` method where the original dataset object is modified unexpectedly. It's intended to show the violation of immutability expectations.

```python
import numpy as np
import xarray as xr

nz = 11
ds = xr.Dataset({
    "y": ("z", np.random.rand(nz)),
    "lev": ("z", np.arange(nz) * 10),
})

# This should not modify ds, but it does
ds2 = ds.swap_dims(z="lev").rename_dims(lev="z").reset_index("lev").reset_coords()
ds2.swap_dims(z='lev')  # This modifies ds2['lev'] unexpectedly

# ds2['lev'] now has dimension 'lev' instead of 'z'
```

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

### Check Build Output (Unix/macOS)

Source: https://github.com/zilliztech/claude-context/blob/master/packages/chrome-extension/CONTRIBUTING.md

Verify the contents of the build output directory on Unix-like systems.

```bash
# Check build output (Unix/macOS)
ls -la dist/
```

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

### Create Global Config .env File

Source: https://github.com/zilliztech/claude-context/blob/master/docs/getting-started/environment-variables.md

Create a global configuration file in ~/.context/.env to store environment variables for embedding providers and Milvus connection details. Ensure you replace placeholder values with your actual API keys and endpoints.

```bash
mkdir -p ~/.context
cat > ~/.context/.env << 'EOF'
EMBEDDING_PROVIDER=OpenAI
OPENAI_API_KEY=sk-your-openai-api-key
EMBEDDING_MODEL=text-embedding-3-small
MILVUS_ADDRESS=your-zilliz-cloud-public-endpoint
MILVUS_TOKEN=your-zilliz-cloud-api-key
EOF
```

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

### Django ORM Filter with __iso_year

Source: https://github.com/zilliztech/claude-context/blob/master/evaluation/case_study/django_14170/README.md

This example demonstrates the incorrect filtering behavior when using `__iso_year` with Django's ORM. The generated SQL uses `BETWEEN` which is inappropriate for ISO years, unlike the expected `EXTRACT` function.

```python
# This should use EXTRACT('isoyear' FROM ...) but incorrectly uses BETWEEN
DTModel.objects.filter(start_date__iso_year=2020)
# Generates: WHERE "start_date" BETWEEN 2020-01-01 AND 2020-12-31
# Should be: WHERE EXTRACT('isoyear' FROM "start_date") = 2020
```

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

### Build All Packages

Source: https://github.com/zilliztech/claude-context/blob/master/CONTRIBUTING.md

Build all packages within the project. This command compiles the source code for all modules.

```bash
pnpm build
```

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

### Verify Bundle Size (Windows PowerShell)

Source: https://github.com/zilliztech/claude-context/blob/master/packages/chrome-extension/CONTRIBUTING.md

Check the size of the built distribution files on Windows using PowerShell, displaying size in kilobytes.

```powershell
Get-ChildItem dist/ | Select-Object Name, @{Name="Size";Expression={[math]::Round($_.Length/1KB,2)}} | Format-Table -AutoSize
```

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

### Build Specific Packages (Bash)

Source: https://github.com/zilliztech/claude-context/blob/master/README.md

Commands to build individual packages within the project.

```bash
pnpm build:core
pnpm build:vscode
pnpm build:mcp
```

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

### Index and Search Codebase with Claude Context Core

Source: https://github.com/zilliztech/claude-context/blob/master/packages/core/README.md

Initialize the embedding provider and vector database, create a context instance, index a local codebase, and then perform a semantic search. Progress is logged during indexing.

```typescript
import {
  Context,
  OpenAIEmbedding,
  MilvusVectorDatabase
} from '@zilliz/claude-context-core';

// Initialize embedding provider
const embedding = new OpenAIEmbedding({
  apiKey: process.env.OPENAI_API_KEY || 'your-openai-api-key',
  model: 'text-embedding-3-small'
});

// Initialize vector database
const vectorDatabase = new MilvusVectorDatabase({
  address: process.env.MILVUS_ADDRESS || 'localhost:19530',
  token: process.env.MILVUS_TOKEN || ''
});

// Create context instance
const context = new Context({
  embedding,
  vectorDatabase
});

// Index a codebase
const stats = await context.indexCodebase('./my-project', (progress) => {
  console.log(`${progress.phase} - ${progress.percentage}%`);
});

console.log(`Indexed ${stats.indexedFiles} files with ${stats.totalChunks} chunks`);

// Search the codebase
const results = await context.semanticSearch(
  './my-project',
  'function that handles user authentication',
  5
);

results.forEach(result => {
  console.log(`${result.relativePath}:${result.startLine}-${result.endLine}`);
  console.log(`Score: ${result.score}`);
  console.log(result.content);
});
```

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

### Configure OpenAI Codex CLI for Claude Context MCP

Source: https://github.com/zilliztech/claude-context/blob/master/packages/mcp/README.md

Add this TOML configuration to your `~/.codex/config.toml` file to integrate the Claude Context MCP server with OpenAI Codex CLI. This includes environment variables and an optional startup timeout override.

```toml
# IMPORTANT: the top-level key is `mcp_servers` rather than `mcpServers`.
[mcp_servers.claude-context]
command = "npx"
args = ["@zilliz/claude-context-mcp@latest"]
env = { "OPENAI_API_KEY" = "your-openai-api-key", "MILVUS_TOKEN" = "your-zilliz-cloud-api-key" }
# Optional: override the default 10s startup timeout
startup_timeout_ms = 20000

```

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

### Configure Claude Context for Gemini CLI

Source: https://github.com/zilliztech/claude-context/blob/master/docs/getting-started/quick-start.md

Add this JSON configuration to `~/.gemini/settings.json` to set up Claude Context with Gemini CLI. Replace placeholders with your actual API keys.

```json
{
  "mcpServers": {
    "claude-context": {
      "command": "npx",
      "args": ["@zilliz/claude-context-mcp@latest"],
      "env": {
        "OPENAI_API_KEY": "your-openai-api-key",
        "MILVUS_TOKEN": "your-zilliz-cloud-api-key"
      }
    }
  }
}
```

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

### Cursor/Claude Desktop Development Mode Configuration

Source: https://github.com/zilliztech/claude-context/blob/master/packages/mcp/CONTRIBUTING.md

JSON configuration for setting up the MCP server in development mode for Cursor or Claude Desktop. Ensure to replace placeholder API keys and paths.

```json
{
  "mcpServers": {
    "claude-context-local": {
      "command": "node",
      "args": ["PATH_TO_CLAUDECONTEXT/packages/mcp/dist/index.js"],
      "env": {
        "OPENAI_API_KEY": "sk-your-openai-api-key",
        "MILVUS_TOKEN": "your-zilliz-cloud-api-key"
      }
    }
  }
}
```

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

### Build Chrome Extension Commands

Source: https://github.com/zilliztech/claude-context/blob/master/packages/chrome-extension/CONTRIBUTING.md

Use these commands to build, watch, clean, lint, type check, and generate icons for the Chrome extension.

```bash
pnpm build:chrome
```

```bash
pnpm dev:chrome
```

```bash
pnpm clean
```

```bash
pnpm lint
```

```bash
pnpm typecheck
```

```bash
pnpm prebuild
```

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

### Build Applications with Claude Context Core Package

Source: https://github.com/zilliztech/claude-context/blob/master/README.md

Use this snippet to initialize the Claude Context core package for indexing codebases and performing semantic searches. Ensure you have set up your OpenAI API key and Milvus connection details.

```typescript
import { Context, MilvusVectorDatabase, OpenAIEmbedding } from '@zilliz/claude-context-core';

// Initialize embedding provider
const embedding = new OpenAIEmbedding({
    apiKey: process.env.OPENAI_API_KEY || 'your-openai-api-key',
    model: 'text-embedding-3-small'
});

// Initialize vector database
const vectorDatabase = new MilvusVectorDatabase({
    address: process.env.MILVUS_ADDRESS || 'your-zilliz-cloud-public-endpoint',
    token: process.env.MILVUS_TOKEN || 'your-zilliz-cloud-api-key'
});

// Create context instance
const context = new Context({
    embedding,
    vectorDatabase
});

// Index your codebase with progress tracking
const stats = await context.indexCodebase('./your-project', (progress) => {
    console.log(`${progress.phase} - ${progress.percentage}%`);
});
console.log(`Indexed ${stats.indexedFiles} files, ${stats.totalChunks} chunks`);

// Perform semantic search
const results = await context.semanticSearch('./your-project', 'vector database operations', 5);
results.forEach(result => {
    console.log(`File: ${result.relativePath}:${result.startLine}-${result.endLine}`);
    console.log(`Score: ${(result.score * 100).toFixed(2)}%`);
    console.log(`Content: ${result.content.substring(0, 100)}...`);
});
```

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

### Configure Claude Context for OpenAI Codex CLI

Source: https://github.com/zilliztech/claude-context/blob/master/docs/getting-started/quick-start.md

Add this TOML configuration to `~/.codex/config.toml` to set up Claude Context with OpenAI Codex CLI. Replace placeholders with your actual API keys.

```toml
# IMPORTANT: the top-level key is `mcp_servers` rather than `mcpServers`.
[mcp_servers.claude-context]
command = "npx"
args = ["@zilliz/claude-context-mcp@latest"]
env = { "OPENAI_API_KEY" = "your-openai-api-key", "MILVUS_TOKEN" = "your-zilliz-cloud-api-key" }
# Optional: override the default 10s startup timeout
startup_timeout_ms = 20000
```

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

### Configure OpenAI Embeddings

Source: https://github.com/zilliztech/claude-context/blob/master/packages/mcp/README.md

Set up the OpenAI embedding provider by providing your API key and optionally specifying the model and a custom API base URL.

```bash
# Required: Your OpenAI API key
OPENAI_API_KEY=sk-your-openai-api-key

# Optional: Specify embedding model (default: text-embedding-3-small)
EMBEDDING_MODEL=text-embedding-3-small

# Optional: Custom API base URL (for Azure OpenAI or other compatible services)
OPENAI_BASE_URL=https://api.openai.com/v1
```

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

### Configure Gemini Embeddings

Source: https://github.com/zilliztech/claude-context/blob/master/packages/mcp/README.md

Set up Gemini embeddings with your API key, optionally specifying the model and a custom API base URL.

```bash
# Required: Your Gemini API key
GEMINI_API_KEY=your-gemini-api-key

# Optional: Specify embedding model (default: gemini-embedding-001; supports gemini-embedding-2)
EMBEDDING_MODEL=gemini-embedding-001

# Optional: Custom API base URL (for custom endpoints)
GEMINI_BASE_URL=https://generativelanguage.googleapis.com/v1beta
```

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

### Run Other MCP Clients

Source: https://github.com/zilliztech/claude-context/blob/master/docs/getting-started/quick-start.md

Integrate with any MCP-compatible client by running the claude-context MCP server using npx. This method uses stdio transport and follows the standard MCP protocol.

```bash
npx @zilliz/claude-context-mcp@latest
```

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

### Configure Zilliz Cloud Vector Database

Source: https://github.com/zilliztech/claude-context/blob/master/packages/mcp/README.md

Provide your Zilliz Cloud API token and optionally configure a timeout for Milvus collection limit pre-checks.

```bash
MILVUS_TOKEN=your-zilliz-cloud-api-key

# Optional: increase timeout for Milvus collection-limit pre-check on slow clusters (default: 15000)
MILVUS_COLLECTION_LIMIT_CHECK_TIMEOUT_MS=30000
```

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

### Configure Cursor for Claude Context

Source: https://github.com/zilliztech/claude-context/blob/master/README.md

Paste this JSON configuration into your Cursor `~/.cursor/mcp.json` file or a project-specific `.cursor/mcp.json` to add Claude Context. Ensure all placeholder values are replaced with your actual credentials.

```json
{
  "mcpServers": {
    "claude-context": {
      "command": "npx",
      "args": ["-y", "@zilliz/claude-context-mcp@latest"],
      "env": {
        "OPENAI_API_KEY": "your-openai-api-key",
        "MILVUS_ADDRESS": "your-zilliz-cloud-public-endpoint",
        "MILVUS_TOKEN": "your-zilliz-cloud-api-key"
      }
    }
  }
}
```

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

### Set Custom Extensions via Environment Variable

Source: https://github.com/zilliztech/claude-context/blob/master/docs/dive-deep/file-inclusion-rules.md

Configure additional file extensions to be indexed by setting the `CUSTOM_EXTENSIONS` environment variable.

```bash
export CUSTOM_EXTENSIONS=".vue,.svelte,.astro"
```

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

### Configure Qwen Code for Claude Context

Source: https://github.com/zilliztech/claude-context/blob/master/README.md

Add this JSON configuration to your `~/.qwen/settings.json` file to enable Claude Context integration with Qwen Code. Replace the placeholder values with your specific API keys and endpoint.

```json
{
  "mcpServers": {
    "claude-context": {
      "command": "npx",
      "args": ["@zilliz/claude-context-mcp@latest"],
      "env": {
        "OPENAI_API_KEY": "your-openai-api-key",
        "MILVUS_ADDRESS": "your-zilliz-cloud-public-endpoint",
        "MILVUS_TOKEN": "your-zilliz-cloud-api-key"
      }
    }
  }
}
```

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

### Configure Void for Claude Context MCP

Source: https://github.com/zilliztech/claude-context/blob/master/packages/mcp/README.md

Add this JSON configuration to your Void MCP settings to integrate the Claude Context MCP server. It includes the command, arguments, and environment variables for authentication and connection.

```json
{
  "mcpServers": {
    "code-context": {
      "command": "npx",
      "args": ["-y", "@zilliz/claude-context-mcp@latest"],
      "env": {
        "OPENAI_API_KEY": "your-openai-api-key",
        "MILVUS_ADDRESS": "your-zilliz-cloud-public-endpoint",
        "MILVUS_TOKEN": "your-zilliz-cloud-api-key"
      }
    }
  }
}

```

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

### Check Build Output (Windows PowerShell)

Source: https://github.com/zilliztech/claude-context/blob/master/packages/chrome-extension/CONTRIBUTING.md

Verify the contents of the build output directory on Windows PowerShell.

```powershell
# Check build output (Windows PowerShell)
Get-ChildItem dist/ | Format-Table -AutoSize
```

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

### Configure Milvus for Local Deployment

Source: https://github.com/zilliztech/claude-context/blob/master/docs/troubleshooting/faq.md

Set these environment variables to configure the Milvus vector database for a local deployment. Ensure Milvus is running via Docker Compose.

```bash
export MILVUS_ADDRESS=127.0.0.1:19530
export MILVUS_TOKEN=your-optional-token
```

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

### Configure Windsurf MCP Server

Source: https://github.com/zilliztech/claude-context/blob/master/docs/getting-started/quick-start.md

Integrate this JSON configuration into your Windsurf MCP settings to set up the Claude Context MCP server. Remember to use your valid API keys and endpoints.

```json
{
  "mcpServers": {
    "claude-context": {
      "command": "npx",
      "args": ["-y", "@zilliz/claude-context-mcp@latest"],
      "env": {
        "OPENAI_API_KEY": "your-openai-api-key",
        "MILVUS_TOKEN": "your-zilliz-cloud-api-key"
      }
    }
  }
}
```

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

### Configure Void MCP Server

Source: https://github.com/zilliztech/claude-context/blob/master/docs/getting-started/quick-start.md

Add this configuration to your Void MCP settings to enable the Claude Context MCP server. Ensure you replace placeholder API keys and endpoints with your actual credentials.

```json
{
  "mcpServers": {
    "code-context": {
      "command": "npx",
      "args": ["-y", "@zilliz/claude-context-mcp@latest"],
      "env": {
        "OPENAI_API_KEY": "your-openai-api-key",
        "MILVUS_ADDRESS": "your-zilliz-cloud-public-endpoint",
        "MILVUS_TOKEN": "your-zilliz-cloud-api-key"
      }
    }
  }
}
```

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

### Configure Augment MCP Server

Source: https://github.com/zilliztech/claude-context/blob/master/README.md

Add this configuration to the `mcpServers` array in `augment.advanced.settings.json` to set up the Claude Context MCP server within Augment.

```json
"augment.advanced": {
  "mcpServers": [
    {
      "name": "claude-context",
      "command": "npx",
      "args": ["-y", "@zilliz/claude-context-mcp@latest"],
      "env": {
        "OPENAI_API_KEY": "your-openai-api-key",
        "MILVUS_ADDRESS": "your-zilliz-cloud-public-endpoint",
        "MILVUS_TOKEN": "your-zilliz-cloud-api-key"
      }
    }
  ]
}
```

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

### Configure Claude Context for Cursor (OpenAI)

Source: https://github.com/zilliztech/claude-context/blob/master/docs/getting-started/quick-start.md

Add this JSON configuration to your Cursor `~/.cursor/mcp.json` for OpenAI embedding provider. Replace placeholders with your actual API keys.

```json
{
  "mcpServers": {
    "claude-context": {
      "command": "npx",
      "args": ["-y", "@zilliz/claude-context-mcp@latest"],
      "env": {
        "EMBEDDING_PROVIDER": "OpenAI",
        "OPENAI_API_KEY": "your-openai-api-key",
        "MILVUS_TOKEN": "your-zilliz-cloud-api-key"
      }
    }
  }
}
```

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

### Grep Method: Reading File Snippet

Source: https://github.com/zilliztech/claude-context/blob/master/evaluation/case_study/pydata_xarray_6938/README.md

This log demonstrates the grep method reading a specific range of lines from `xarray/core/dataarray.py` to inspect the `swap_dims` method documentation. This is part of the process to narrow down the relevant code sections.

```log
🔧 read_file('xarray/core/dataarray.py', offset=2000, limit=40)
⚙️ Result: Read lines 2001-2040 showing rename method documentation
```

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

### Watch Mode for Core Package Development

Source: https://github.com/zilliztech/claude-context/blob/master/packages/core/CONTRIBUTING.md

Run this command to enable watch mode for the core package, which automatically rebuilds on file changes. This is useful during active development.

```bash
pnpm dev:core
```

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

### Chrome Extension Project Structure

Source: https://github.com/zilliztech/claude-context/blob/master/packages/chrome-extension/CONTRIBUTING.md

Overview of the source directory structure for the Chrome extension.

```tree
src/
├── content.ts              # Content script for GitHub integration
├── background.ts           # Background service worker
├── options.ts             # Options/settings page
├── options.html           # Options page HTML
├── styles.css             # Extension styles
├── manifest.json          # Extension manifest
├── config/
│   └── milvusConfig.ts    # Milvus configuration
├── milvus/
│   └── chromeMilvusAdapter.ts  # Browser Milvus adapter
├── storage/
│   └── indexedRepoManager.ts   # Repository management
├── stubs/                 # Browser compatibility stubs
└── icons/                 # Extension icons
```

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

### Semantic Search: Initial Code Search

Source: https://github.com/zilliztech/claude-context/blob/master/evaluation/case_study/pydata_xarray_6938/README.md

This log shows the initial step of using semantic search to find occurrences of `.swap_dims()` within the Xarray codebase. It specifies the repository path and the search query.

```log
🔧 search_code('.swap_dims()')
⚙️ Arguments: {'path': 'repo__pydata__xarray', 'query': '.swap_dims()', 'limit': 10}
   Result: Found 10 results for query: ".swap_dims()" in codebase
   1. Code snippet (python) [repo__pydata__xarray]
      Location: xarray/tests/test_formatting_html.py:67-76
   2. Code snippet (python) [repo__pydata__xarray]
      Location: xarray/tests/test_dataset.py:2926-2979
```

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

### Configure Zilliz Cloud Connection

Source: https://github.com/zilliztech/claude-context/blob/master/examples/basic-usage/README.md

Configure Milvus connection details for Zilliz Cloud. Set the MILVUS_ADDRESS to your Public Endpoint and MILVUS_TOKEN to your Zilliz token.

```bash
export MILVUS_ADDRESS="https://your-cluster.zillizcloud.com"
export MILVUS_TOKEN="your-zilliz-token"
```

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

### Configure Claude Context for Cursor (Ollama)

Source: https://github.com/zilliztech/claude-context/blob/master/docs/getting-started/quick-start.md

Add this JSON configuration to your Cursor `~/.cursor/mcp.json` for Ollama embedding provider. Ensure Ollama is running and replace placeholders.

```json
{
  "mcpServers": {
    "claude-context": {
      "command": "npx",
      "args": ["-y", "@zilliz/claude-context-mcp@latest"],
      "env": {
        "EMBEDDING_PROVIDER": "Ollama",
        "EMBEDDING_MODEL": "nomic-embed-text",
        "OLLAMA_HOST": "http://127.0.0.1:11434",
        "MILVUS_TOKEN": "your-zilliz-cloud-api-key"
      }
    }
  }
}
```

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

### Configure Zilliz Cloud Connection

Source: https://github.com/zilliztech/claude-context/blob/master/packages/vscode-extension/README.md

Set environment variables for your Zilliz Cloud public endpoint and API key. These are required for connecting to your Milvus vector database.

```bash
MILVUS_ADDRESS=your-zilliz-cloud-public-endpoint
MILVUS_TOKEN=your-zilliz-cloud-api-key
```

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

### Simplified MCP Configuration for Claude Context

Source: https://github.com/zilliztech/claude-context/blob/master/docs/getting-started/environment-variables.md

Configure Claude Context within MCP using environment variables passed directly via the command line for Claude-compatible environments. For other environments like Cursor or Windsurf, use a JSON configuration file to specify the command and arguments.

```bash
claude mcp add claude-context \
  -e OPENAI_API_KEY=sk-your-openai-api-key \
  -e MILVUS_ADDRESS=your-zilliz-cloud-public-endpoint \
  -e MILVUS_TOKEN=your-zilliz-cloud-api-key \
  -- npx @zilliz/claude-context-mcp@latest
```

```json
{
  "mcpServers": {
    "claude-context": {
      "command": "npx",
      "args": ["-y", "@zilliz/claude-context-mcp@latest"]
    }
  }
}
```

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

### Claude Code CLI Usage

Source: https://github.com/zilliztech/claude-context/blob/master/README.md

Basic commands for interacting with Claude Code. Use `claude` to open the application, followed by commands to index, check status, and search your codebase.

```bash
cd your-project-directory
claude
```

```bash
Index this codebase
```

```bash
Check the indexing status
```

```bash
Find functions that handle user authentication
```

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

### Configure Augment MCP Server

Source: https://github.com/zilliztech/claude-context/blob/master/docs/getting-started/quick-start.md

Add the server configuration to the `mcpServers` array in the `augment.advanced` object within your settings.json file.

```json
"augment.advanced": {
  "mcpServers": [
    {
      "name": "claude-context",
      "command": "npx",
      "args": ["-y", "@zilliz/claude-context-mcp@latest"]
    }
  ]
}
```

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

### Configure Claude Context for Cursor (Gemini)

Source: https://github.com/zilliztech/claude-context/blob/master/docs/getting-started/quick-start.md

Add this JSON configuration to your Cursor `~/.cursor/mcp.json` for Gemini embedding provider. Replace placeholders with your actual API keys.

```json
{
  "mcpServers": {
    "claude-context": {
      "command": "npx",
      "args": ["-y", "@zilliz/claude-context-mcp@latest"],
      "env": {
        "EMBEDDING_PROVIDER": "Gemini",
        "GEMINI_API_KEY": "your-gemini-api-key",
        "MILVUS_TOKEN": "your-zilliz-cloud-api-key"
      }
    }
  }
}
```
Tokens:
Raw