Codegen SDK Documentation

The Codegen MCP server is an implementation of the Model Context Protocol (MCP) that provides tools and resources for interacting with the Codegen platform APIs. It is designed to enable AI agents to manage development workflows and access Codegen services directly through an MCP-compatible interface.

Dependencies:

• fastmcp

Codegen sandboxes use a custom Docker image designed as a comprehensive development environment. It is optimized for full-stack and polyglot development, providing pre-configured runtimes, package managers, and system utilities.

Core Runtimes:

• Python 3.13 (via uv)
• Node.js 22.14.0 (managed via NVM)

Key Capabilities:

• Multi-Language Support: Out-of-the-box support for Python and Node.js.
• Package Management: Includes uv (Python), npm, yarn, pnpm (Node.js), and corepack.
• Development Tools: git, gh (GitHub CLI), ripgrep, fd-find, tmux, and nginx.
• Remote Access: Pre-configured SSH server for remote development.

The Notion integration allows Codegen agents to bridge the gap between planning and implementation by performing the following tasks:

• Access workspace documentation: Read and analyze existing documentation, specs, and project requirements.
• Update pages and databases: Modify content, add comments, and keep documentation current.
• Sync development information: Connect planning documents directly to code implementation.
• Generate knowledge base content: Create new documentation based on development work and insights.

While you can use the RESTful API directly, Codegen provides tools for easier integration:

• Python SDK: Provides a pythonic interface for creating and managing agents.
• CLI Tool: A command-line interface for performing common API operations.

The GitHub integration enables Codegen to perform a full spectrum of agentic coding workflows, including:

• Create and manage pull requests: Generate, update, and manage PRs with detailed descriptions and context.
• Automated code reviews and feedback: Provide intelligent code analysis and suggestions.
• Run checks and CI/CD workflows: Execute automated testing and deployment processes.
• Sync repository changes: Keep repositories up-to-date and coordinate between branches.

The codegen CLI provides a terminal-optimized version of the web UI and API capabilities, including:

• Agent Management: List agent runs, check status, and view execution logs.
• Work Retrieval: Pull branches and code changes created by agents to your local environment.
• Agent Creation: Trigger new agent runs using custom prompts.
• Claude Code Integration: Run Claude Code with OpenTelemetry monitoring and logging.
• Organization Management: Switch between organizations and configure repositories.
• Telemetry: Access traces and monitoring via MCP injection and web UI integration.

Codegen is SOC 2 Type I & II certified and undergoes regular penetration testing to ensure security, privacy, and compliance.

For detailed information, you can visit:

• Security Overview: https://codegen.com/security
• Trust Center: https://trust.codegen.com/ (for SOC-2 documents and pen test results)

Enterprise customers deploying Codegen on-premises receive additional support and features:

• Priority Support: Dedicated channels and faster response times.
• Custom Configuration: Tailored deployment plans for specific organizational requirements.
• Security Integration: Compatibility with existing security tools and compliance policies.
• Multi-Region Support: High-availability configurations across multiple clusters.
• Air-gapped/Offline Support: Supported via special configuration.
• Engineering Access: Direct access to the engineering team for deployment assistance and optimization reviews.

The Sentry integration allows Codegen agents to interact with your error tracking data to accelerate production problem resolution.

Capabilities

• Automated root cause analysis: Analyzing error patterns and stack traces.
• Error investigation: Deep diving into error contexts, user impact, and code changes.
• Issue prioritization: Identifying critical errors.
• Performance monitoring: Analyzing application bottlenecks.
• Release tracking: Connecting errors to specific deployments.
• Team coordination: Assigning issues based on expertise.

How Agents Use Sentry

Agents use the integration to examine stack traces, generate resolution recommendations, update issue statuses, and link error fixes to GitHub PRs and code changes.

The Figma integration enables several design-to-code workflows:

• Access design specifications: Read design files, components, and detailed specifications.
• Extract design assets: Pull images, icons, and visual elements for implementation.
• Convert designs to code: Transform design mockups into functional frontend code.
• Sync design changes: Stay updated with design iterations and modifications.