databricks-industry-solutions

Add tools to your agent and grant required permissions in databricks.yml. Use when: (1) Adding MCP servers, Genie spaces, vector search, or UC functions to agent, (2) Permission errors at runtime, (3) User says 'add tool', 'connect to', 'grant permission', (4) Configuring databricks.yml resources.

Add memory capabilities to your agent. Use when: (1) User asks about 'memory', 'state', 'remember', 'conversation history', (2) Want to persist conversations or user preferences, (3) Adding checkpointing or long-term storage.

Deploy agent to Databricks Apps using DAB (Databricks Asset Bundles). Use when: (1) User says 'deploy', 'push to databricks', or 'bundle deploy', (2) 'App already exists' error occurs, (3) Need to bind/unbind existing apps, (4) Debugging deployed apps, (5) Querying deployed app endpoints.

Configure Lakebase for agent memory storage. Use when: (1) Adding memory capabilities to the agent, (2) 'Failed to connect to Lakebase' errors, (3) Permission errors on checkpoint/store tables, (4) User says 'lakebase', 'memory setup', or 'add memory'.

Migrate an MLflow ResponsesAgent from Databricks Model Serving to Databricks Apps. Use when: (1) User wants to migrate from Model Serving to Apps, (2) User has a ResponsesAgent with predict()/predict_stream() methods, (3) User wants to convert to @invoke/@stream decorators.

Set up authentication for the ADME OSDU connector — get an Azure bearer token and write dev_config.json for live testing.

Mechanical audit of the ADME OSDU connector — implementation correctness, test coverage, artifacts completeness, and security. Produces a scored report.

Run the AdmeOsduLakeflowConnect pytest suite, diagnose failures, and fix the connector or simulator until everything passes. Branches on mode={simulate|live}.

Validate that credentials for the ADME OSDU connector authenticate successfully against the real ADME API.

Validate that ADME domain tables (wellbore, reservoir, rock_and_fluid) correctly implement CDC incremental sync — offset tracking, watermark filtering, and pagination termination.

Based on a compound, get its safety info. Assess compound safety profile including toxicity, hazard classifications, and regulatory information using PubChem and PubMed.

Based on a compound, get its ADME and other properties. Assess compounds for ADME (Absorption, Distribution, Metabolism, Excretion) characteristics, chemical properties, and drug-likeness using PubChem. Use when the user wants to evaluate compound suitability as a lead candidate. Triggers include requests like "assess drug-likeness for [compound]", "evaluate ADME for [CID]", "ADME assessment", "check Lipinski rules for [compound]", "molecular properties of [drug]", or "is [compound] a good lead candidate". Accepts PubChem CIDs, compound names, SMILES, or InChI as input.

Based on a target, get its associated drugs. Identify small molecule hits for therapeutic targets by querying Open Targets and PubChem. Use when the user provides a gene symbol (e.g., EGFR, BRAF, KRAS) or protein name and wants to find known compounds, drugs, or chemical matter with activity against that target. Returns compound identifiers, names, bioactivity data, clinical trial phases, mechanism of action summaries, and supporting literature with PubMed links. Triggers include requests like "find hits for [target]", "what compounds bind [gene]", "identify drugs targeting [protein]", "small molecules for [target]", or "hit identification for [gene symbol]".

Based on a disease, identify therapeutic targets. Use this skill when users ask to find drug targets, therapeutic targets, or druggable genes for any disease or condition. Triggers include queries like "find druggable targets for [disease]", "what are therapeutic targets for [disease]", "identify drug targets for [condition]", or any request to discover targetable genes/proteins for a disease. The skill uses Open Targets Platform for disease-target associations, clinical precedence, and tractability data, combined with PubMed for supporting mechanistic evidence. Outputs a ranked table of top 10 targets with evidence summaries.

Audit npm and Python dependencies across all Databricks Apps in a workspace. Use when checking for malicious packages, generating dependency inventories, or investigating supply chain risks. Triggers on dependency audit, npm audit, package audit, supply chain, malicious package, dependency inventory.

Kickstart Many Models Forecasting (MMF) projects on Databricks — explore data, profile series, configure clusters, run forecasting pipelines, and evaluate results.

Development guidance for the dbxmetagen project -- AI-powered metadata generation for Databricks Unity Catalog. Use when working on dbxmetagen source code, tests, DDL generation, metadata review workflow, semantic layer, FK prediction, ontology, the FastAPI+React dashboard app, or deployment via Databricks Asset Bundles. Triggers on dbxmetagen, metadata generation, DDL, review workflow, knowledge base, knowledge graph, Genie builder, metric views.

Quick reference for apx toolkit commands and MCP tools for building Databricks Apps