Data Lineage Viz
Visualise SQL dependencies right inside VS Code. Ask @lineage in Copilot Chat to explore your lineage graph in natural language — or browse interactively with search, trace, and schema overview.
Import from .dacpac files or connect directly to SQL Server, Azure SQL, Fabric Data Warehouse, or Synapse Dedicated SQL Pool.
Get started
- Run Data Lineage: Open Wizard (
Ctrl+Shift+P).
- Pick a
.dacpac file — or Connect to Database via the MSSQL extension.
- Select schemas and click Visualize.
No data? Click Try with demo data to explore the AdventureWorks sample.
AI lineage with @lineage
Use @lineage in GitHub Copilot Chat to explore dependencies in natural language. The assistant answers from your loaded data model — never from general knowledge.
@lineage trace from Sales.SalesOrderDetail upstream to the source tables
@lineage how is sales calculated — show me the lineage in the app
@lineage which objects are hubs with the most connections?
The assistant can trace dependencies, build bookmarked graph views, and analyse column mappings and SQL logic from the available metadata.
@lineage has two modes of operation:
- Discovery (chat) — the default. Most lineage questions are answered directly in chat, including multi-object dependency questions like "trace upstream from X two levels" or "what feeds Y". Each visited node gets its own heading with business meaning, technical execution, formulas in math fences, column-rename tables, and ⚠️ data-quality flags inline.
- Structured walkthrough (graph render + column tracing) — when you ask for a graph rendered in the side panel, request a column trace, or open a scope too large for chat, the assistant first shows you the planned scope (nodes, schemas, depth) and asks for confirmation. Once approved, it walks the graph hop-by-hop and renders the result with origin / transformer / terminal nodes coloured.
Requires GitHub Copilot. Tools activate automatically when a graph is loaded.
Features
@lineage AI — trace lineage, analyse column mappings, explain SQL logic, and build bookmarked graph views from the loaded model.- Search & trace — autocomplete object lookup, upstream / downstream tracing, shortest-path between nodes.
- Graph analysis — islands, hubs, orphans, circular dependencies, longest chains.
- Schema overview — large graphs auto-summarise at schema level; double-click to drill in.
- SQL preview — click any node to view DDL with syntax highlighting; full-text search across procedure / view bodies.
- Multiple sources — SSDT and SDK-style
.dacpac, live database connections, external tables, virtual external refs (OPENROWSET, cross-DB, CETAS).
- Projects & views — save connections, schema selections, and named filter states for one-click reopen.
- Table profiling — on-demand column statistics for live databases (null %, distinct, min / max, AVG, STDEV).
- Export — Draw.io diagram generation.
For the full feature catalogue, settings, and customisation paths see docs/FEATURES.md.
Limitations
The extension covers intra-database DDL only. The following are out of scope:
- External ingestion pipelines — ADF, SSIS, Spark, Fabric Dataflow, or any ETL/ELT process that writes into the database from an external source. Target tables appear as leaves; the upstream pipeline does not.
- Cross-database / cross-server flow — surfaced only when the SQL body uses a fully qualified three- or four-part name; the remote side appears as an external-reference node, not a fully traced sub-graph.
- Dynamic SQL —
EXEC(@sql) and sp_executesql cannot be analysed statically.
- Unqualified references — references without a schema prefix are excluded.
FAQ
Do I need a .dacpac file?
No — connect directly to a database. If you prefer a .dacpac, extract one from Visual Studio, SSMS, Azure Data Studio, or the Fabric portal. See Microsoft's documentation.
Why are some dependencies missing?
Dynamic SQL cannot be analysed statically. Only compile-time dependencies are detected.
Why are unqualified references excluded?
Unqualified names depend on the caller's default schema, which the catalog does not record reliably. Schema-qualified names are the only way to be sure.
Documentation
Contributing
Bug reports welcome. For custom features, fork and extend under the MIT license. See CONTRIBUTING.md.
MIT License · Christian Wagner · GitHub · Developed with Claude Code, Codex and Gemini
