adf-graph
MCP server that builds a queryable dependency graph from ADF pipeline artifacts
Versions
1.10.1latest1.10.01.9.01.8.01.7.0+ show 33 moreshow less
1.6.31.6.21.6.11.6.01.5.01.4.11.4.01.3.11.3.01.2.11.2.01.1.31.1.21.1.11.1.01.0.11.0.00.9.30.9.20.9.10.9.00.8.30.8.20.8.10.8.00.7.00.6.10.6.00.5.00.3.00.2.00.1.10.1.0Tools 42
graph_stats Returns aggregate statistics about the ADF dependency graph (node/edge counts by type, build time, staleness).
graph_export Export the full graph (all nodes and edges) as a single JSON payload. Designed for visualization tools that need the complete topology.
graph_find_consumers Find all pipeline activities that consume a given dataset, table, stored procedure, or Dataverse entity.
graph_describe_pipeline Describe a pipeline: summary, activities, full detail, or resolved (inlines parameter values for child pipeline calls, detects CDC patterns). Optionally filter to a single named activity.
graph_describe_entity Describe a Dataverse entity: metadata, attributes, and pipeline consumers. At
graph_impact_analysis Analyse which nodes are affected if a given artifact changes. Traverses upstream, downstream, or both.
graph_data_lineage Trace data lineage for a Dataverse entity or staging table. Optionally filter to a single attribute/column.
graph_find_paths Find all dependency paths between two nodes in the graph.
graph_find_orchestrators Find root orchestrator pipelines that own a given pipeline. Returns full ancestry chains with depth.
graph_search Flexible search across the graph: node names, activity SQL, FetchXML, stored procedure names/parameters, and ExecutePipeline parameter values. Supports filters for activity type, node type, target entity, and pipeline scope.
graph_trace_connection Trace the full connection chain from a pipeline
graph_diff_pipeline Compare a pipeline
graph_diff_environments Compare pipelines across two environments. Returns added/removed/changed pipelines with summary-level diffs.
graph_diff_staging Compare staged pipeline changes against the deployed version. Auto-detects staging and deployed environments from config, or accepts explicit environment names.
graph_cross_env_artifact Compare a single artifact across all registered environments. Shows per-environment metadata with field-level diffs to spot configuration inconsistencies (e.g. different serviceUri across factories).
graph_validate Run graph-wide validation: broken references, empty-default parameters without suppliers, unused datasets, orphaned nodes, cross-org Dataverse URI mismatches. Returns errors and warnings.
graph_validate_pipeline Validate dest_query column aliases against Dataverse entity schema. Checks that each SQL alias maps to a valid entity attribute, flags invalid columns, and whitelists system attributes.
graph_validate_statuscode Validate CASE WHEN values for statuscode/statecode columns in dest_query against Dataverse OptionSet metadata. Checks that integer values map to valid OptionSet options.
graph_find_bad_columns Bulk audit: scan all pipelines for dest_query parameters and report every column alias that does not match a Dataverse entity attribute.
graph_validate_staging_columns Validate source_query SELECT columns against staging table DDL. Detects column name mismatches that cause ADF auto-mapping failures at runtime. Warns when Copy activities use zero explicit mappings.
graph_ignore_null_values_audit Scan all Copy activities writing to Dataverse and flag those with ignoreNullValues absent or false. This dangerous default causes NULL source columns to overwrite existing Dataverse values.
graph_sp_body Return the full SQL body of a stored procedure. Use this when you need the actual SQL content, not just metadata.
graph_staging_dependencies Map shared staging table usage across pipelines. Shows which pipelines read/write each table, flags shared tables where concurrent execution risks data corruption, and detects TRUNCATE TABLE patterns.
graph_entity_coverage Show all pipelines writing to a Dataverse entity with per-pipeline column lists. Highlights column differences across pipelines to detect mapping inconsistencies.
graph_trace_parameters Trace parameter flow through ExecutePipeline chains from a root pipeline. Maps each parameter from source to sink and flags dead-ends: parameters with empty/null defaults that no caller supplies a value for.
graph_parameter_trace Trace parameter values from parent to child pipelines. For a given pipeline, show what each caller supplies for each parameter via ExecutePipeline activities. Flags dead-end parameters with no supplier.
graph_deploy_readiness Pre-flight check: walks the full dependency tree of a pipeline and reports what artifacts are present, stub (referenced but no file), or missing in the target environment. Also flags parameters with empty/null defaults that no parent supplies. Optionally compares linked service configuration against another environment.
graph_list_environments List all configured environments with their paths, default status, and graph statistics (node/edge counts, last build time, staleness).
graph_add_overlay Add an overlay path (directory or file) to an environment. The overlay
graph_remove_overlay Remove a runtime overlay from an environment. Config-based overlays cannot be removed via this tool.
graph_list_overlays List all overlays (config-based and runtime) for an environment.
graph_add_environment Register a new ephemeral environment pointing to an ADF artifact directory. Lost on server restart. Cannot collide with config-based environment names.
graph_remove_environment Remove a runtime environment. Config-based environments cannot be removed via this tool.
graph_generate_scope Generate a scope manifest by walking orchestrator pipeline trees. Collects all reachable pipelines, stored procedures, tables, and datasets. Optionally detects orphan pipelines in a specified ADF folder.
graph_filter_chain Extract and display all WHERE/filter conditions across the pipeline chain for a given entity or table. Shows the complete filter path from source through staging to destination.
graph_cdc_analysis Analyse CDC (Change Data Capture) pipeline configuration. Shows source CDC tables, staging tables (current/historical/pending), the full filter chain from source through staging to Dataverse, escape hatch conditions, and detects configuration gaps.
graph_staging_population Cross-reference staging tables in a pipeline
graph_describe_stored_procedure Return structured facts about a stored procedure: parameters, tables read/written, callers, confidence, and column mappings. Avoids reading .sql files directly.
graph_describe_table Return a table
graph_describe_trigger Return trigger schedules, associated pipelines, and runtime state. If no trigger name is given, lists all triggers.
graph_describe_integration_runtime Return integration runtime type, compute config, and which linked services use it. If no IR name is given, lists all IRs.
graph_environment_config Return linked service endpoint configuration for a deployment target (e.g.
Permissions 2
filesystem low env_vars low