Integrations and MCP

Connect Zencoder with your tools and workflows through multiple integration paths

Overview

Zencoder connects to external tools through three integration paths:

Native Integrations

Jira integration built into the IDE plugin

Chrome Extension

Send context from 20+ browser-based tools to your IDE

Model Context Protocol

Open standard for connecting agents to databases, APIs, and custom tools

Built-in Native Integrations

Add Jira Integration

The Jira integration lets you reference Jira tickets directly in agent chat using @Jira. Agents can read issue details (summary, description, acceptance criteria) to understand what to build or fix.

1. Access Integrations

   1. Click the three dots menu (⋮) at the top right corner
   2. Select Integrations from the dropdown menu
2. Connect Your Jira Account

   1. In the Integrations screen, find the Jira card
   2. Click Connect to begin the integration process
3. Authorize Access

   1. Follow the prompts to authorize Zencoder to access your Jira instance
   2. You’ll be redirected to Jira to approve the connection
   3. Once approved, you’ll be returned to Zencoder
4. Start Using Jira in Zencoder

   Once connected, you can reference Jira tickets directly in your Zencoder chat by using the @Jira mention.

Chrome Extension

The Chrome extension adds a Zencoder button to supported browser-based tools. Clicking it captures the current page context (error stack traces, PR descriptions, issue details) and sends it to your IDE.

1. Install the Chrome Extension

   Visit the Chrome Web Store and add the Zencoder extension to your browser.

2. Use with Supported Tools

   When browsing supported tools, click the Zencoder button to send context to your IDE. No authentication needed!

Supported Tools

Project Management

• Jira
• Linear
• Trello
• Monday
• Wrike
• Asana
• ClickUp
• Shortcut
• Azure DevOps

Code Repositories

• GitHub
• GitLab (including self-hosted instances)

Error Monitoring

• Sentry
• Datadog
• BugSnag
• Rollbar
• SonarQube
• Snyk
• Grafana/Loki
• LogRocket

CI/CD

• CircleCI
• BuildKite

Frontend Debugging

• Chrome DevTools (with a dedicated tab)

Model Context Protocol (MCP)

MCP is an open standard for connecting LLMs to external data sources and tools. Zencoder acts as an MCP client that connects to MCP servers—lightweight connectors that expose specific capabilities (database queries, API calls, browser automation, etc.).

If you need to connect to a system without a native integration, you can use an existing MCP server or build your own.

MCP Protocol Support

Supported connection protocols:

• Standard stdio communication: For local MCP servers running as subprocesses
• Streamable HTTP connections: For remote MCP servers using HTTP-based communication
• OAuth2 authentication: For secure authentication with third-party services requiring OAuth2 flows

Platform Availability:

• JetBrains IDEs: Full support for all MCP protocols (stdio, streamable HTTP, OAuth2) available from version 2.13 (end of August 2025)
• VS Code: Currently supports stdio connections. Streamable HTTP and OAuth2 support coming soon!

Managing Agent Tools and MCP Servers

MCP servers can be managed through the Agent Tools menu in the IDE, or by editing configuration files directly.

Agent Tools UI (Suggested)

1. Access Agent Tools menu

   1. Click the three dots menu (⋮) at the top right corner
   2. Select Agent Tools to access the tools management interface
2. Navigating Agent Tools Categories

   At the top of the Agent Tools screen, you’ll find tabs that help you filter and navigate between different types of tools:

   • All: Shows all available tools (both Zencoder-provided and custom)

   • Zencoder: Displays prebuilt tools created by Zencoder

   • Custom: Shows your custom tools and MCPs that you’ve installed

   • MCP Library: Provides a selection of ready-made MCP servers that you can install with a single click and minimal setup

     
3. Browse available tools

   The Agent Tools interface displays all available tools based on the category you’ve selected, including pre-built MCP servers from the MCP Library and any custom tools you’ve added.

4. Install from MCP Library

   To install a pre-built MCP server:

   1. Click on the MCP Library tab
   2. Browse or search through the available MCP servers

   Installed servers will have an “Installed” label next to their name, making it easy to see which MCPs you’ve already added.

   3. Click on the MCP server you want to install
   4. Fill in the required configuration details in the form
   5. In some cases, you may need to select an installation method (such as Docker or local installation)
   6. Click the Install button at the top of the form to complete the installation
5. Add a custom MCP server

   To add your own custom MCP server:

   1. Click the Add Custom MCP button
   2. Enter a name for your MCP and fill in the configuration details
   3. Click the Install button at the top to add your custom MCP
6. Manage existing MCP servers

   To manage an existing MCP server:

   1. Click on the MCP server in your list of installed servers
   2. You will see its configuration details and name
   3. You can modify these settings as needed
   4. To completely remove the MCP server, click the Uninstall button
7. Apply and test your MCP server

   After adding an MCP server:

   1. Make sure Coding Agent is enabled
   2. Test it with a relevant prompt. For example, if you added the time MCP server, try: “What’s the time in Zagreb right now?”

   Important: Different MCP servers require different command-line tools to be installed on your system:

   • uvx/uv: Many MCP servers use the uvx or uv package manager for Python
   • Docker: Some MCPs require Docker to be installed and running
   • Node.js/npm: Certain MCPs depend on Node.js and npm
   • Language-specific tools: Depending on the MCP, you might need specific language runtimes (Python, Ruby, etc.)

   Ensure these prerequisites are installed on your system before using the corresponding MCP servers. Check the MCP’s documentation for specific requirements.

Manual config (VS Code)

1. Go to Zencoder settings

   Open … menu in Zencoder chat and select Settings from the dropdown menu.

2. Find MCP Servers section

   Scroll down to the bottom of the settings page until you find the MCP Servers part and click on “Edit in settings.json”.

3. Locate the MCP configuration

   In the settings.json file, find the MCP servers configuration section that looks like this:

   "zencoder.mcpServers": {
   
   },
   
4. Add an MCP server

   Let’s add a simple MCP server for getting time information. Add the following configuration:

   "zencoder.mcpServers": {
     "time": {
       "command": "uvx",
       "args": ["mcp-server-time"]
     }
   },
   
5. Save and test

   Save the settings.json file. Now go back to your chat, make sure Coding Agent is enabled and try a prompt like: “What’s the time in Zagreb right now?”

   Important: Ensure you have the required command-line tools installed on your system. In this example, you need the uvx command available in your PATH.

Manual config (JetBrains)

1. Go to Zencoder settings

   Open JetBrains settings by going to File > Settings (or JetBrains IDE > Settings on macOS).

2. Navigate to MCP Servers

   In the settings window, scroll down to Tools and expand it if needed. Find Zencoder and expand it until you see MCP Servers, then click on it.

3. Add MCP server configuration

   You’ll see instructions showing the format:

   {
       "mcp_server_name": {
          "command": "...",
          "args": [...],
          ...
       }
   }
   
4. Add the time server

   Paste the following configuration into the editor:

   {
     "time": {
       "command": "uvx",
       "args": ["mcp-server-time"]
     }
   }
   
5. Apply and test

   Click Apply to save the changes, then reopen your project. Go back to your chat, make sure Coding Agent is enabled and try a prompt like: “What’s the time in Zagreb right now?”

   Important: Ensure you have the required command-line tools installed on your system. In this example, you need the uvx command available in your PATH.