Bigeye MCP Server

An MCP (Model Context Protocol) server that provides tools for interacting with the Bigeye Data Observability platform.

- Docker(Docker Desktop or Docker Engine)

cp .env.example .env # Edit .env with your values

{ "mcpServers": { "bigeye": { "command": "/absolute/path/to/mcp-wrapper.sh" } } }

- BIGEYE_DEBUG— Set totruefor verbose debug logging (default:false)
- BIGEYE_TELEMETRY— Set tofalseto disable anonymous usage telemetry (default:true). SeeTelemetry.

{ "mcpServers": { "bigeye": { "command": "/absolute/path/to/mcp-wrapper.sh" } } }

{ "mcpServers": { "bigeye": { "command": "docker", "args": [ "run", "--rm", "-i", "-e", "BIGEYE_API_KEY=your_api_key_here", "-e", "BIGEYE_BASE_URL=https://app.bigeye.com", "-e", "BIGEYE_WORKSPACE_ID=your_workspace_id_here", "-e", "BIGEYE_DEBUG=false", "bigeye-mcp-server:latest" ] } } }

Thebigeye-mcp.shscript manages the long-lived container:

./bigeye-mcp.sh start # Start the container ./bigeye-mcp.sh stop # Stop the container ./bigeye-mcp.sh restart # Restart the container ./bigeye-mcp.sh status # Show container status ./bigeye-mcp.sh logs # Follow container logs ./bigeye-mcp.sh rebuild # Rebuild image and recreate container ./bigeye-mcp.sh clean # Remove container and volumes

To connect to multiple Bigeye instances (e.g. demo and production), create separate environment files and compose overrides:
-

Create environment files for each instance. Note that the compose overrides expectprefixedvariable names (BIGEYE_DEMO_/BIGEYE_APP_):

BIGEYE_DEMO_API_KEY=your_demo_api_key BIGEYE_DEMO_WORKSPACE_ID=your_demo_workspace_id BIGEYE_DEBUG=false

BIGEYE_APP_API_KEY=your_app_api_key BIGEYE_APP_WORKSPACE_ID=your_app_workspace_id BIGEYE_DEBUG=false

Use the environment-specific compose overrides:

# Demo docker compose -f docker-compose.yml -f docker-compose.demo.yml --env-file .env.demo up -d bigeye-mcp-demo # Production docker compose -f docker-compose.yml -f docker-compose.app.yml --env-file .env.app up -d bigeye-mcp-app

{ "mcpServers": { "bigeye-demo": { "command": "docker", "args": [ "run", "--rm", "-i", "-e", "BIGEYE_API_KEY=your_demo_key", "-e", "BIGEYE_BASE_URL=https://demo.bigeye.com", "-e", "BIGEYE_WORKSPACE_ID=your_demo_workspace_id", "bigeye-mcp-server:latest" ] }, "bigeye-app": { "command": "docker", "args": [ "run", "--rm", "-i", "-e", "BIGEYE_API_KEY=your_app_key", "-e", "BIGEYE_BASE_URL=https://app.bigeye.com", "-e", "BIGEYE_WORKSPACE_ID=your_app_workspace_id", "bigeye-mcp-server:latest" ] } } }

- list_issues— List data quality issues across the workspace (filter by status, schema, orassignee_ids)
- get_current_user— Get the authenticated user (id, email, name, workspaces); pair withlist_issues(assignee_ids=[id])to find issues assigned to you
- get_issue— Get full details for a single issue by its internal ID
- search_issues— Find issues by their display name/number (e.g. "10921")
- list_related_issues— List issues related to a given issue via lineage
- list_table_issues— List data quality issues for a specific table by name
- update_issue— Update an issue's status, priority, or add a timeline message
- create_incident— Create an incident by merging related issues
- delete_incident_members— Remove issues from an incident
- get_resolution_steps— Get recommended resolution steps for an issue

- list_table_metrics— List all metrics (monitors) configured on a table
- list_table_level_metrics— List metric types that are table-level (vs column-level)
- create_metric— Create a new metric (monitor) on a table with validation, enum mapping, and column-type compatibility checks
- get_table_profile— Get data profile report including column statistics and distribution
- create_profile_job— Queue a new data profiling job for a table
- get_profile_job_status— Check the status of a profiling job

- list_data_classes— List data classification categories (e.g., "Email Address", "US SSN") with sensitivity levels
- get_scan_findings— Get column-level classification scan results showing where sensitive data was detected

- search_schemas— Search the data catalog for schemas by name (returns ids + warehouse)
- search_tables— Search the data catalog for tables by name (returns ids, schema + warehouse)
- search_columns— Search the data catalog for columns by name (returns ids, type + parent table)

- get_lineage_graph— Get the full lineage graph (upstream/downstream/both) from a starting node
- get_lineage_node— Get details for a specific lineage node
- list_lineage_node_issues— List issues for a lineage node by its node ID
- search_lineage_nodes— Find lineage node IDs by path pattern (e.g. "WAREHOUSE/SCHEMA/TABLE")
- lineage_explore_catalog— Explore tables in Bigeye's catalog
- lineage_delete_node— Delete a custom lineage node

- get_upstream_root_causes— Analyze upstream lineage to identify root causes of issues
- get_downstream_impact— Analyze downstream impact of issues at a lineage node
- get_issue_lineage_trace— Trace a data quality issue end-to-end through lineage
- list_report_upstream_issues— List upstream issues affecting a BI report or dashboard

- lineage_track_data_access— Track data assets accessed by an AI agent
- lineage_commit_agent— Commit tracked data access to Bigeye's lineage graph
- lineage_get_tracking_status— Get the current status of lineage tracking
- lineage_clear_tracked_assets— Clear all tracked data assets without committing
- lineage_cleanup_agent_edges— Clean up old lineage edges for the AI agent

- list_dimensions— List all data-quality dimensions with their metric type mappings
- get_dimension— Get full details for a single dimension by ID
- create_dimension— Create a new Data Dimension
- update_dimension— Update a dimension's name or description
- delete_dimension— Delete a Data Dimension
- get_table_dimension_coverage— Analyze dimension coverage gaps for a table (recommended for "what monitoring is missing?")
- get_column_dimension_coverage— Analyze dimension coverage for specific columns in a table

- list_tags— List or search workspace tags
- create_tag— Create a new tag with optional color
- update_tag— Update a tag's name or color
- delete_tag— Delete a tag
- tag_entity— Apply a tag to any entity (metric, table, column, etc.)
- untag_entity— Remove a tag from an entity
- list_entity_tags— List all tags on a specific entity

- get_health_status— Check the health and connectivity of the Bigeye API
- list_resources— List all available MCP resources
- list_data_sources— List all data sources/warehouses connected to Bigeye

- bigeye://auth/status— Current authentication status
- bigeye://health— API health status
- bigeye://config— Current server configuration
- bigeye://issues— All issues from the configured workspace
- bigeye://issues/active— Active issues with filtering
- bigeye://issues/recent— Recently resolved or updated issues

- authentication_flow— Guide for setting up authentication
- check_connection_info— Guide for verifying API connection
- merge_issues_example— Examples for merging issues
- lineage_analysis_examples— Examples for lineage analysis

The server includes comprehensive lineage tracking for AI agents. Use thelineage_track_data_accesstool to record data assets accessed during an agent session, thenlineage_commit_agentto persist them to Bigeye's lineage graph. See the tool descriptions above for full details.
- Install Python 3.12+
- Create a virtual environment:

python -m venv venv source venv/bin/activate

export BIGEYE_API_KEY="your_api_key" export BIGEYE_BASE_URL="https://app.bigeye.com" export BIGEYE_WORKSPACE_ID="your_workspace_id"

# Run basic container tests ./scripts/test.sh # Run basic tests + MCP protocol tests ./scripts/test.sh --mcp # Run MCP protocol tests standalone (with optional --debug) ./scripts/test-mcp-protocol.sh

Seetests/README.mdfor details on the test suite.

An MCP server for the Dynatrace observability platform.