MCP-Allure

MCP-Allure is an MCP server that reads Allure HTML test reports and transforms their contents into structured, LLM-friendly JSON formats. It connects local or generated Allure reporting directories directly to AI assistants operating via the Model Context Protocol. Software developers, QA automation engineers, and test leads use this server to automate the extraction of complex test metrics without needing to manually parse bulky HTML or raw log artifacts. MCP-Allure exposes specific test execution details, including test suite hierarchies, individual test case parameters, execution status, runtime timestamps, failure steps, and attachments. By organizing test data into clean JSON schemas, it enables language models to comprehend test outputs, perform automated failure triage, generate test health summaries, and suggest remediation steps directly within development workflows.

Category: Developer Tools & Code Intelligence

Tags: allure, automation, qa, reports, testing

Visit MCP-Allure

How to install and configure MCP-Allure

1. Ensure Python, uv, and the mcp package are installed on your environment. 2. Clone or locate the repository locally containing mcp-allure-server.py. 3. Open your MCP client configuration file (such as Claude Desktop or Cursor configuration). 4. Add mcp-allure-server to your mcpServers definition using the uv runner: json { "mcpServers": { "mcp-allure-server": { "command": "uv", "args": [ "run", "--with", "mcp[cli]", "mcp", "run", "/path/to/mcp-allure/mcp-allure-server.py" ] } } } 5. Replace the script path with the absolute path to your local mcp-allure-server.py file and restart your MCP client.

What you can do with MCP-Allure

• Analyze failing test steps and stack traces from an Allure HTML report directory to identify root causes. * Generate automated executive summaries of test execution runs across multiple test suites for pull request reports. * Inspect individual test case parameters, execution durations, and severity levels to detect slow or flaky test cases. * Extract attached logs and failure artifacts directly into the context window for AI-driven code debugging suggestions. * Automate documentation and QA audit logs from local Allure report folders using conversational prompts.

Key facts

• https://github.com/crisschan/mcp-allure
• Developer Tools & Code Intelligence, Monitoring & Observability
• allure, automation, qa, reports, testing

Part of MCP Servers

What is MCP-Allure?

MCP-Allure is a Model Context Protocol server that parses Allure test report directories. It transforms standard Allure test report data into structured JSON objects designed for processing and interpretation by large language models.

What tools does MCP-Allure provide?

The server provides the get_allure_report tool. This tool accepts a directory path (report_dir) pointing to an Allure HTML report and returns formatted JSON containing test suites, test cases, execution statuses, timestamps, labels, parameters, and step attachments.

Which MCP clients work with MCP-Allure?

MCP-Allure works with any client that implements the Model Context Protocol. This includes Claude Desktop, Cursor, and any other MCP-compatible agents or command-line developer tools that support local server execution via standard I/O.

How do I run MCP-Allure locally?

MCP-Allure is configured by adding its execution command to your client configuration using uv. The command runs the mcp CLI with the path to the mcp-allure-server.py script file located on your local machine.