SmartBear

SmartBear

An MCP server exposing SmartBear APIs to AI assistants for seamless access to testing tools.

17
Stars
16
Forks
14
Releases

Overview

An MCP server that exposes SmartBear APIs to AI assistants, enabling seamless access to SmartBear's ecosystem of testing and monitoring tools. Built around the Model Context Protocol (MCP) open standard, it securely connects external data sources and tools so AI workflows can query testing data, analyze performance metrics, and manage test automation directly from conversations. The server is distributed as the npm package @smartbear/mcp and can be started with API keys or per-product auth tokens (tokens are optional and can be removed from configuration if not used). It supports access to BugSnag, Reflect, Swagger (Portal and Studio), PactFlow/Contract Testing, Pact Broker, QMetry, Zephyr, and Collaborator, with prompts and guides available for each integration. Configuration is flexible: BugSnag can be narrowed to a single project using a project API key; per-product base URLs and tokens can be supplied via environment variables or a .vscode/mcp.json file. Quick setup options include the MCP: Add server command in VS Code, Claude Desktop configuration for launching with npx, and comprehensive documentation. Local development and CONTRIBUTING guidelines are provided to assist contributors.

Details

Owner
SmartBear
Language
TypeScript
License
MIT License
Updated
2025-12-07

Features

Natural language access to SmartBear APIs

Provides AI assistants with a natural language interface to query testing data, analyze performance metrics, and manage test automation across SmartBear products.

Unified access to core SmartBear tools

Supports BugSnag, Test Hub, Swagger (Portal and Studio), PactFlow/Contract Testing, Pact Broker, QMetry, Zephyr, and Collaborator through a single MCP server.

Easy installation via npm package

Distributed as the @smartbear/mcp npm package for straightforward installation and integration into development workflows.

Token-based authentication per product

Server starts with API keys or per-product auth tokens; tokens are optional and can be omitted if a product is not in use.

Per-product configuration and optional BugSnag scoping

Supports per-product base URLs and tokens; BugSnag can be narrowed to a single project using a project API key.

VS Code Copilot quick setup

Provides a quick setup path via the MCP: Add server… command to configure the MCP server in VS Code.

Claude Desktop integration

Includes configuration guidance to launch the MCP server from Claude Desktop using npx.

Documentation and Local Development

Offers full documentation, a CONTRIBUTING guide, and local development instructions to facilitate adoption and contribution.

Audience

AI developersIntegrate SmartBear tools with AI assistants to access testing data and automation capabilities.
QA teamsQuery testing data, monitor metrics, and manage test automation via MCP-based AI workflows.
DevOps engineersEmbed SmartBear tools into CI/CD AI assistants for operational visibility and automation management.

Tags

MCPSmartBearAIOpen standardIntegrationBugSnagReflectSwaggerPactFlowPact BrokerQMetryZephyrCollaboratorCLINode.jsNPM