lucius-mcp 0.2.2

Allure TestOps MCP Server

Lucius MCP Server

A Model Context Protocol (MCP) server for Allure TestOps, built with FastMCP and Starlette.

🚀 Features

• FastMCP Integration: Leverages the FastMCP framework for efficient tool and resource management.
• Starlette Mounting: Mounted as a Starlette application for robust HTTP handling and easy extension.
• Structured Logging: JSON-formatted logging with automatic secret masking (powered by src/utils/logger.py).
• Global Error Handling: User-friendly "Agent Hint" error responses optimized for LLM consumption (powered by src/utils/error.py).
• Type Safety: Fully typed codebase checked with mypy --strict.
• Quality Assurance: Linting and formatting with ruff.

⚙️ Configuration

The server can be configured using environment variables or a .env file.

Variable	Description	Default
ALLURE_ENDPOINT	Allure TestOps Base URL	https://demo.testops.cloud
ALLURE_PROJECT_ID	Default Project ID	None
ALLURE_API_TOKEN	Allure API Token	None
LOG_LEVEL	Logging level	INFO
HOST	Host to bind the server to	127.0.0.1
PORT	Port to bind the server to	8000
MCP_MODE	Running mode: http or stdio	stdio

🛠️ Installation

This project uses uv for dependency management.

1. Install uv (if not already installed):

   curl -LsSf https://astral.sh/uv/install.sh | sh
   

2. Clone the repository:

   git clone https://github.com/lucius-mcp/lucius-mcp.git
   cd lucius-mcp
   

3. Install dependencies:

   uv sync
   

🏃 Usage

Add to Claude Code (MCP)

1. Add the MCP:

   claude mcp add testops-mcp --transport stdio \
     --env ALLURE_ENDPOINT=https://your-testops.example \
     --env ALLURE_PROJECT_ID=123 \
     --env ALLURE_API_TOKEN=your_token \
     --env MCP_MODE=stdio \
     -- uvx --from lucius-mcp start
   

Running via Claude Desktop (One-Click Install)

The easiest way to use Lucius in Claude Desktop is via the .mcpb bundle:

1. Download the latest lucius-mcp-x.y.z.mcpb from the GitHub Releases.
2. Open the file with Claude Desktop (macOS or Windows).
3. Approve the installation in the dialog.
4. Configure your Allure TestOps credentials in the provided fields.

Running via Stdio

For integration with MCP clients (like Claude Code) using standard input/output.

uv run start

Or

MCP_MODE=stdio uv run start

Running via HTTP

Starts the server with hot-reloading enabled (default port: 8000).

MCP_MODE=http uv run start

Or customizing host and port:

HOST=0.0.0.0 PORT=9000 uv run start

Local Workflow Testing (act)

To run GitHub Actions locally using nektos/act:

1. Install act:

   brew install act
   

2. Run a workflow:

   # Run a specific workflow (e.g., CI)
   act -W .github/workflows/pr-quality-gate.yml 
   

   Note: The --artifact-server-path flag is required to handle artifact uploads/downloads locally. Artifacts will be stored in .artifacts/.

📦 Packaging (MCPB)

Lucius supports the MCP Bundle (MCPB) format for easy distribution.

Building for Claude Desktop

To build a .mcpb bundle locally:

1. Install Node.js (v20+ recommended).
2. Install the mcpb CLI:

   npm install -g @anthropic-ai/mcpb
   

3. Run the build script:

   ./deployment/scripts/build-mcpb.sh
   

   The versioned bundles will be available in the dist/ directory:
   • lucius-mcp-<version>-uv.mcpb
   • lucius-mcp-<version>-python.mcpb

Manifests

Bundle manifests live in deployment/mcpb/:

• manifest.uv.json
• manifest.python.json

Validation

Validate each manifest against the code:

python deployment/scripts/validate_mcpb.py uv
python deployment/scripts/validate_mcpb.py python

UV Runtime

The uv bundle uses the uv runtime type, which means:

• Dependencies are defined in pyproject.toml and uv.lock.
• Claude Desktop will automatically manage the Python environment and dependencies for the user.
• No local Python installation is required for the end-user.

🧪 Testing

Run the test suite using pytest:

uv run pytest

End-to-End (E2E) Tests

E2E tests verify the integration with a real Allure TestOps instance. They are isolated in tests/e2e/ and strictly separated from unit/integration tests.

Prerequisites:

1. Sandbox Environment: Access to a non-production Allure TestOps instance.
2. Configuration: Create .env.test from .env.test.example:

   cp .env.test.example .env.test
   

Running E2E Tests:

# Load environment variables from .env.test
uv run --env-file .env.test pytest tests/e2e/

Troubleshooting E2E Failures:

• 401 Unauthorized: Check ALLURE_API_TOKEN. It might be expired.
• 403 Forbidden: Ensure ALLURE_PROJECT_ID exists and the user has Write access.
• Connection Errors: Verify ALLURE_ENDPOINT is reachable and uses HTTPS.
• Flaky Tests: Tests use unique IDs to avoid collisions, but network issues can occur. Rerunning usually fixes transient issues.

🛠️ Development

Regenerating API Client

To maintain spec-fidelity while keeping the client lightweight, we use a 2-step process:

1. Filter Spec: scripts/filter_openapi.py reduces the massive OpenAPI spec to only the essential controllers (Test Cases, Shared Steps, Projects).
2. Generate Client: openapi-generator-cli builds the client from the filtered spec.

• Generated Client (src/client/generated/): Auto-generated ApiClient, API controllers, and Pydantic v2 models.
• Client Facade (src/client/client.py): AllureClient wrapper that handles authentication, error mapping, and helper methods.
• Model Facade (src/client/models/): Re-exports generated models for simplified import paths and logical grouping.

To regenerate the client after updating the spec:

./scripts/generate_testops_api_client.sh

Note: Do not manually edit files in src/client/generated/.

🧹 Quality Checks

Formatting:

uv run ruff format .

Linting:

uv run ruff check src/

Type Checking:

uv run mypy --strict src/

Release process

1. Bump version in pyproject.toml.
2. Run uv sync --all-extras to update dependencies.
3. Write release notes in CHANGELOG.md.
4. Commit changes.
5. Create a tag with the new version.
6. Push commits and tags.