vibe-coding-workshop-template

Scaffold new Databricks AppKit applications using the Databricks CLI and Agent Skills. Creates blank or plugin-enabled AppKit projects (Lakebase, Analytics, Genie, Files). Use when asked to create a Databricks app, scaffold an AppKit project, bootstrap a new app, or set up a full-stack TypeScript Databricks application. Triggers on "create app", "new app", "scaffold", "AppKit", "databricks app", "blank app", "bootstrap app", "init app". To add a plugin to an existing app, use the 04-appkit-plugin-add skill instead.

Entry-point navigator for Databricks AppKit + Lakebase development. Routes tasks to the correct skill based on keyword detection: scaffolding, plugin integration, feature building, or deployment. This is a routing skill -- it does not generate code. It directs the agent to the correct specialized skill. Use this skill as the starting point for any AppKit-related task. Triggers on "AppKit", "Lakebase app", "Databricks app", "build app", "deploy app", "scaffold app", "add plugin", "AppKit project", "agent endpoint", "wire agent", "serving plugin", "agent app proxy", "chat history", "feedback".

Build full-stack UI and backend features on a Databricks AppKit project from a PRD or feature spec. Covers SQL query design, type generation, React frontend with AppKit UI components, backend plugin wiring, and distinctive visual design. Use when asked to implement a UI, build features from a PRD, create pages or dashboards, add components, develop the frontend/backend of an existing AppKit app, build with mock data, or use static data for visual prototyping. Triggers on "build UI", "implement PRD", "create dashboard", "add page", "build features", "implement design", "create components", "build app from PRD", "develop frontend", "mock data", "static data", "two-phase data".

Deploy a Databricks AppKit application to Databricks Apps. Covers config validation, build verification, deployment, UI verification, error diagnosis with fix loop, and workspace app limit handling. Use when asked to deploy an AppKit app, push to production, ship the app, or troubleshoot a failed deploy. Triggers on "deploy app", "push to databricks", "ship app", "deploy appkit", "databricks apps deploy", "fix deploy error", "app won't start".

Add plugins to an existing Databricks AppKit project. Covers Lakebase (PostgreSQL), Analytics (SQL queries + dashboards), Genie (natural language AI/BI), Files (UC Volumes), and Serving (Model Serving / Agent endpoints). Guides through plugin registration, environment variables, app.yaml configuration, and frontend integration. Use when asked to add a plugin to an existing app, integrate Lakebase, add analytics, connect a Genie space, enable file uploads, add a serving endpoint, or extend an AppKit project with new capabilities. Triggers on "add plugin", "add lakebase", "add analytics", "add genie", "add files plugin", "add serving", "add agent endpoint", "integrate postgres", "add database", "add dashboards", "add file browser", "extend app", "connect genie space", "model serving plugin". Not for creating a Lakebase project or managing Lakebase via CLI -- use the databricks-lakebase agent skill for that.

Wire a Lakebase PostgreSQL backend into an existing AppKit project. Covers database schema design from a PRD, idempotent DDL, Express API routes with mock fallback, React data hooks, and local testing. PRD-independent patterns that apply to any AppKit + Lakebase app. Use after registering the Lakebase plugin via 04-appkit-plugin-add. Triggers on "wire lakebase", "lakebase backend", "CRUD API", "lakebase tables", "DDL", "database schema design", "useLakebaseData", "mock fallback", "ConnectionStatus", "replace mock data", "connect frontend to backend", "API-backed data", "replace static data with database".

Wire a Databricks Model Serving or Agent endpoint into an existing AppKit project using the Serving plugin. Covers endpoint registration, app.yaml resource binding, streaming and invoke React hooks, conversation state management, agent response mapping, and server-side proxy patterns. PRD-independent patterns that apply to any AppKit + Serving app. Use after registering the Serving plugin via 04-appkit-plugin-add. Triggers on "wire agent", "agent endpoint", "serving plugin", "agent UI", "chat interface", "connect agent", "model serving", "useServingStream", "useServingInvoke", "agent chat", "wire serving", "serving wiring".

Wire an AppKit frontend (Databricks App) to a separate Agent App backend (Databricks App hosting mlflow.genai.agent_server or ResponsesAgent) with OBO token forwarding via x-forwarded-access-token, app-to-app service principal auth, dual-format SSE streaming, and the /api/chat contract. Use this skill when the agent is deployed as its own Databricks App — not as a Model Serving endpoint. Triggers on "agent app backend", "two apps", "appkit plus agent app", "proxy agent app", "agent-backend resource", "x-forwarded-access-token", "migrate agent to apps", "agent app proxy".

Add persistent chat history to an AppKit application using the Lakebase plugin and the Serving plugin (both AppKit-native, no Vercel AI SDK). Covers database schema, session auth middleware, a server-side proxy route that streams from `AppKit.serving("agent").asUser(req).stream()` while persisting assistant messages and capturing MLflow `trace_id`, the history API routes, a history sidebar, and ephemeral-mode fallback. Use when asked to persist conversations, add chat history, save messages, or build a conversation sidebar. Triggers on "chat history", "save conversations", "persistent chat", "conversation sidebar", "message storage", "save messages", "ephemeral chat".

Add user feedback (thumbs up/down) to an AppKit chat application, linked to MLflow assessments via the Databricks Assessments REST API. Covers the Vote table, feedback API routes (with AppKit-native auth via `getExecutionContext().client.config.authenticate()`), MLflow trace integration, and feedback UI components. Use when asked to add feedback, thumbs up/down, ratings, or link user judgments to MLflow traces. Triggers on "feedback", "thumbs up", "thumbs down", "rate response", "MLflow assessment", "user rating", "vote on message".

GenAI agent skill navigator for the `genai-agents/` folder. Start here to route agent-building, MLflow GenAI, tracing, Databricks Apps Track A, SDLC, evaluation, deployment, monitoring, prompt iteration, and vibecoding runtime tasks to the correct specialized skill. Uses tiered loading and keyword routing so agents read only the skills needed for the current task.

Session-start behavioral manifest for running this workshop inside Databricks Genie Code. Read FIRST when the client is Genie Code (the coding agent embedded in the Databricks workspace) so you begin knowing how it behaves — surface/page tool-scoping, the three execution paths (runDatabricksCli → SDK → native tools) and the "blocked ≠ impossible, try the next path" discipline, the runDatabricksCli allow-list tiers, bundle-deploy reality (--target dev mandatory, CWD pinned to the page's bundle root, FUSE create-then-validate gap), AppKit/Node reality (apps init --output-dir, no local npm but server-side build, SDK app deploy), agent-skills git-clone install, the Genie-Space deploy tiers, and how to verify a deployed app (3-hop OAuth session). Detection of WHICH client is active lives in vibecoding-state; this skill explains HOW the Genie Code client behaves. Not needed for the IDE+CLI client.

MLflow and ML Model patterns for Databricks including experiment creation, model training, batch inference, and Unity Catalog integration. Use when implementing ML pipelines, training models with Feature Store, or deploying batch inference jobs. Includes 19 non-negotiable rules covering experiment paths, dataset logging, UC model registration, NaN handling, label binarization, feature engineering workflows, and signature-driven preprocessing.

Standard patterns for creating Databricks Metric Views with semantic metadata for Genie and AI/BI. Use when creating metric views, troubleshooting metric view creation errors, validating schema references before deployment, implementing joins (including snowflake schema patterns), or optimizing metric views for Genie natural language queries.

Comprehensive patterns for Databricks Genie Space Export/Import API - JSON schema, serialization format, and programmatic deployment. Use when programmatically creating, exporting, or importing Genie Spaces via REST API, troubleshooting API deployment errors, or implementing CI/CD for Genie Spaces. Includes complete GenieSpaceExport schema, API endpoints (List, Get, Create, Update, Delete), JSON format requirements, ID generation, variable substitution, inventory-driven generation patterns, and production deployment checklists.

Use when deriving a Databricks Agent Spec and Agent Tool Plan from docs/design_prd.md before building Track A agents. Covers web-researched MCP recommendations, Databricks managed MCP choices, dynamic SQL MCP catalog/schema inputs, optional Knowledge Assistant selection, and validation of docs/agent_spec.yaml plus docs/agent_tool_plan.yaml.

Use when starting any new GenAI agent project on Databricks. Covers MLflow 3.x GenAI setup: install dependencies, enable autolog, configure connection pooling, understand model signatures, detect runtime environment. Foundation Step 1. Produces the configured MLflow environment consumed by all subsequent skills.

Use when setting up MLflow experiments, tracing, or UC OTEL trace storage for a GenAI agent. Covers structured experiment paths, tracing decorators, manual spans, tags, connection pooling, and Unity Catalog OTEL storage for SQL-queryable trace retention. Foundation Step 2. Consumes MLflow environment from Step 1.

Use when adding MLflow tracing to a Node/TypeScript GenAI agent. Covers the official `mlflow-tracing` and `mlflow-openai` npm packages: `mlflow.init()`, `tracedOpenAI`, manual `mlflow.trace`/`@mlflow.trace`/`mlflow.withSpan`, nested spans, sessions/users, and when to choose a custom OTLP fallback. Foundation Step 2b. Sibling to the Python tracing skill (Foundation Step 2). Consumes the MLflow environment from Step 1.

Use when connecting any agent to data and external capabilities. Covers Databricks managed MCP servers (Vector Search, Genie, SQL, UC Functions), the databricks-mcp client library, _meta parameters, external MCP servers, code interpreter, non-MCP connections, resource grants, custom retriever schemas, tool cookbook (web search, SQL, file generation, HTTP), runtime guardrails, and end-to-end external MCP examples. Foundation Step 3. Consumed by all agent tracks (A, B, C).

Use when you need central governance, observability, guardrails, or rate limits across LLM endpoints and MCP servers used by agents. Covers creating and configuring Databricks AI Gateway endpoints, enabling usage tracking, inference tables (UC Delta audit), rate limits, and guardrails, plus how deployed agent serving endpoints and Apps consume a gateway URL. Foundation Step 4 — ideal as a governance layer BEFORE Module 7 production deployment. Complements in-code guardrails and MLflow tracing; does NOT replace MLflow trace tables but adds a provider-level audit trail.

Use when creating, syncing, or managing a Databricks Agent Bricks Knowledge Assistant (KA) as a document Q&A source for an agent. Covers source mode selection (UC Files vs Vector Search index), Python SDK lifecycle (create/update/sync/delete), REST API equivalents, readiness polling, and emitting the ka_endpoint_name handoff that downstream agent skills wire as a knowledge_assistant tool. Track-neutral foundation skill. Optional but recommended whenever document Q&A is part of the agent.

Use when you need to version, manage, or load prompts for any GenAI agent on Databricks. Covers how to register prompts in Unity Catalog, set up aliases for safe deployments, link prompts to traces, and load them in production apps — even if you just want "register my prompt and give me a stable URI." Also use when A/B testing prompt versions or connecting prompt changes to evaluation runs. SDLC Step 1.

Use when you need to create, manage, or load evaluation datasets for testing agent quality. Covers the MLflow GenAI data format, persisting benchmarks in Unity Catalog, merging records without duplicates, and validating data before evaluation — even if you just want "give me a dataset I can pass to mlflow.genai.evaluate()." Also use when building benchmarks from production traces or SME labels. SDLC Step 2.

Use when you need to measure agent quality or create scoring criteria for evaluation gates. Covers how to pick and configure scorers — even if you just want "is my agent safe and accurate?" without knowing which MLflow classes to use. Also use when building custom LLM judges, evaluating multi-turn conversations, or setting pass/fail thresholds. SDLC Step 3.

Use when running mlflow.genai.evaluate() to test agent quality before deployment. Covers the predict_fn contract, answer-sheet mode for re-scoring existing outputs, threshold gates, retry wrappers, human feedback sessions, and conversation evaluation — even if you just want "run my benchmarks and tell me pass or fail." Also use when collecting human labels to calibrate automated scorers. SDLC Step 4.

Use when wiring **end-user feedback** (thumbs up/down, ratings, free-form comments) from a deployed app back into MLflow as Assessments on the originating trace. Covers the canonical `mlflow.log_feedback(...)` pattern, trace-id vs `client_request_id` correlation, the streaming case (SSE/WebSocket delivery of `trace_id` after the stream ends), multi-dimensional feedback (rating-per-aspect), update/delete of existing assessments, and analysis queries over collected feedback. Triggers on "collect user feedback", "end user feedback", "thumbs up production", "log_feedback", "trace_id feedback", "feedback API", "feedback route", "user rating production", "feedback assessment". SDLC Step 4c.

Use when you need to log an agent to MLflow, register it in Unity Catalog, or promote a new version as champion. Covers the full model lifecycle — even if you just want "log my agent and make it the production version." Also use when setting up MLFLOW_ACTIVE_MODEL_ID for production trace linking, metric-based promotion gating, or rollback. SDLC Step 5.

Use when deploying an agent to production or setting up CI/CD automation. Covers Databricks Apps deployment, Asset Bundles, service principal permissions, and evaluate-then-promote pipelines — even if you just want "deploy my agent and set up a release gate." Also use for MCP tool connectivity, Supervisor long-running tasks, or production trace linking. SDLC Step 6.

Use when setting up live quality monitoring for a deployed agent. Covers registered scorers with sampling, trace archival to Delta, metric backfill, and periodic human review — even if you just want "monitor my agent's safety score in production." Also use when adding conversation-level judges, tuning sampling rates, or building monitoring dashboards from UC trace data. SDLC Step 7.

Use when prompt quality has plateaued on your evaluation dataset and you need to programmatically improve one or more registered prompts. Covers `mlflow.genai.optimize_prompts()` and `GepaPromptOptimizer`, `predict_fn` contracts, single and multi-prompt optimization, custom scorers, budget controls, registering the optimized output back to UC with a staging alias, and the promote-after-eval loop. SDLC Step 8 — run AFTER Step 4 evaluation reveals specific regressions or unmet guideline checks. Do NOT use for the first pass of an agent (Step 1 Prompt Registry first), or when quality gaps are tool/retrieval problems rather than instruction problems.

Use when building a custom agent with the OpenAI Agents SDK and MLflow ResponsesAgent. Covers Agent class, Runner, @function_tool, handoffs, streaming, ModelConfig, mlflow.models.set_model(), autolog, and manual tracing. Track A Step 2. Produces a customized, MLflow-compatible agent class ready for tools and deployment.

Use when wiring tools into an OpenAI Agents SDK agent: local Python function tools, DatabricksMCPClient connections, and MCPServerSse for MCP servers. Track A Step 3. Builds on shared concepts from F3 (Tools and Data Access).

Use when configuring agent permissions, implementing OBO, or granting access to Databricks resources. Covers App Authorization (service principal), User Authorization (OBO), auth passthrough for Model Serving, and MCP server authentication. Track A Step 4.

Use when implementing stateful agents that need conversation memory. Covers short-term continuity via AsyncDatabricksSession (prescribed for OpenAI Agents SDK), long-term preferences via DatabricksStore, graceful degradation, and thread_id resolution. Track A Step 5. Consumes authenticated client from Step 4.

Use when evaluating an agent's response quality and safety before deployment. Covers running agent-evaluate, evaluation dataset format, built-in judges (relevance, groundedness, safety), interpreting results, and customizing eval datasets. Track A Step 6. Consumes a working agent with tools from Steps 1-5. Produces evaluation results and confidence to deploy.

Use when deploying an agent to Databricks Apps and querying the deployed endpoint. Covers databricks apps create, sync, deploy, OAuth token auth, PAT limitations, curl and Python SDK query methods, and the redeploy workflow. Track A Step 7. Consumes evaluation results from Step 6. Produces a running deployed agent on Databricks Apps.

Scaffold a minimal MCP tool-calling agent with Genie Spaces and deploy it to Databricks Model Serving in 5 steps, following the canonical OpenAI MCP Tool Calling Agent notebook verbatim. Produces a working endpoint testable in AI Playground and consumable by 06-appkit-serving-wiring. No evaluation, memory, or prompt registry — add those later via the existing worker skills. Use when creating a simple agent, scaffolding a new agent, building a quick tool-calling agent, or connecting an agent to Genie Spaces. Triggers on "simple agent", "scaffold agent", "MCP agent", "quick agent", "create agent", "tool calling agent", "Genie agent", "basic agent".

Runtime schema validation patterns for Gold layer merge scripts. Use when creating merge scripts to ensure DataFrame columns match target DDL schemas, validate column mappings before MERGE operations, and catch schema issues before deployment. Includes the DDL-first workflow, validate_merge_schema() helper, explicit column mapping, DDL schema reader, and pre-deployment validation script. Addresses the "three sources of truth" problem by enforcing DDL as runtime truth.

Production-grade patterns for Databricks AI/BI (Lakeview) dashboards. Prevents visualization errors, deployment failures, and maintenance issues through widget-query alignment, number formatting, parameter configuration, monitoring table patterns, chart scale properties, and automated deployment workflows. Includes pivot tables with hierarchy drill-down and ratio metrics, point/choropleth maps, sankey diagrams, waterfall and histogram charts, cross-filtering and drill-through patterns, filter defaultSelection/disallowAll configuration, disaggregated vs aggregated query modes, and complete JSON templates for all widget types.