ART MCP Server

Model Context Protocol (MCP) server for querying BigQuery tables containing ART (Automated Release Tooling Team) build data.

• query_art_builds: Execute custom queries with filtering, field selection, and ordering
• get_build_summary: Get aggregated statistics and metrics about builds
• search_builds_by_package: Find builds containing specific packages
• get_failed_builds: Analyze failed builds for debugging

1. Install dependencies:

pip install -e .

2. Configure BigQuery access:

# Set up authentication
gcloud auth application-default login

# Copy and edit environment file
cp .env.example .env
# Edit .env with your BigQuery project/dataset/table details

3. Run the server:

python tools/bigquery_mcp_server.py

Set these environment variables:

• BIGQUERY_PROJECT_ID: Your GCP project ID
• BIGQUERY_DATASET_ID: BigQuery dataset name
• BIGQUERY_TABLE_ID: BigQuery table name

To make it work with Claude Code:

claude mcp add "bigquery-art" "python" "/Users/asdas/JetBrains/PycharmProjects/art-mcp/tools/bigquery_mcp_server.py"