Skip to content

Latest commit

 

History

History
237 lines (189 loc) · 14.7 KB

File metadata and controls

237 lines (189 loc) · 14.7 KB

Repository Navigator (Quick Reference)

This file is the universal entry point for any AI coding assistant — Cursor, Claude Code, Windsurf, Copilot, Codex, or any agent that reads AGENTS.md.

This is a monorepo with four components: a Databricks AppKit workshop with Lakebase and agent-chat extensions, a standalone GenAI agent course, a Data Product Accelerator with 77 agent skills, and an Agentic Framework for building multi-agent systems.


Project Layout

vibe-coding-workshop-template/          <-- workspace root / agent CWD
├── AGENTS.md                           <-- THIS FILE (root navigator)
├── README.md                           <-- Human-readable project overview
├── QUICKSTART.md                       <-- Quick-start guide (two pathways)
├── PRE-REQUISITES.md                   <-- Workshop prerequisites checklist
├── env.example                         <-- Environment variable template
│
├── apps_lakebase/                      <-- Component 1: Databricks AppKit Workshop
│   ├── Instructions.md                 #   Comprehensive workshop guide
│   ├── prompts/                        #   Numbered prompt files for each workshop step
│   └── skills/                         #   10 agent skills for the full app lifecycle
│       ├── 00-appkit-navigator/        #     Entry-point navigator (read first)
│       ├── 01-appkit-scaffold/         #     Scaffold new AppKit projects
│       ├── 02-appkit-build/            #     Build UI + backend from a PRD
│       ├── 03-appkit-deploy/           #     Deploy to Databricks Apps
│       ├── 04-appkit-plugin-add/       #     Add plugins (Lakebase, Analytics, Genie, Files, Serving)
│       ├── 05-appkit-lakebase-wiring/  #     Wire Lakebase backend (DDL, APIs, frontend hooks)
│       ├── 06-appkit-serving-wiring/   #     Wire Model Serving / Agent endpoint to frontend
│       ├── 06d-appkit-agent-app-proxy/ #     Wire AppKit frontend to a separate Agent App
│       ├── 07-appkit-chat-history/     #     Persist chat conversations in Lakebase
│       └── 08-appkit-feedback/         #     Add thumbs feedback linked to MLflow assessments
│
├── genai-agents/                       <-- Component 2: GenAI Agent Development Course
│   ├── 00-course-orchestrator/         #   Entry-point navigator for the course
│   ├── PROMPT-GUIDE.md                 #   Canonical Track A + AppKit 2-Apps prompt cookbook
│   ├── foundation/                     #   UC, MLflow, tracing, tools, AI Gateway, KA
│   ├── tracks/                         #   Track A custom agent on Databricks Apps
│   └── sdlc/                           #   Prompt registry, evals, deployment, monitoring
│
├── agentic-framework/                  <-- Component 3: Multi-Agent Build Framework
│   ├── agents/                         #   Agent prompts (PRD analyzer, skill scaffolder, etc.)
│   └── skills/
│       └── foundation-model-agent-loop/
│           └── SKILL.md                #   Tool-calling loop with Foundation Models
│
├── presentations/                      <-- Workshop slide deck (Marp, HTML, PDF, PPTX)
│
├── data_product_accelerator/           <-- Component 4: 77 Agent Skills for Data Products
│   ├── AGENTS.md                       #   *** DETAILED SKILL ROUTING TABLE ***
│   ├── QUICKSTART.md                   #   One-prompt-per-stage guide
│   ├── README.md                       #   Accelerator overview
│   ├── context/                        #   Customer schema CSV inputs
│   ├── skills/                         #   77 skills across 12 domains
│   └── docs/                           #   Framework design documentation
│
├── gold_layer_design/                  <-- GENERATED by Gold Design skill
├── src/                                <-- GENERATED by implementation skills
├── plans/                              <-- GENERATED by Planning skill
├── resources/                          <-- GENERATED by Asset Bundle skills
└── databricks.yml                      <-- GENERATED by Asset Bundle skills
Direction Location Examples
Read from (framework) apps_lakebase/, agentic-framework/, data_product_accelerator/ Skills, instructions, agent prompts, docs
Write to (artifacts) Repository root gold_layer_design/, src/, plans/, resources/, databricks.yml

Rule: Generated artifact paths (gold_layer_design/, src/, plans/, resources/, databricks.yml) are relative to the repository root. Never create generated files inside data_product_accelerator/, apps_lakebase/, or agentic-framework/.


Component Routing

MANDATORY: Before starting any task, match keywords below to find the right component. Read the linked file FIRST.

Data Product Accelerator — data_product_accelerator/AGENTS.md

Keywords Action
"medallion architecture", "data product", "lakehouse", "data platform" Read data_product_accelerator/AGENTS.md
"bronze layer", "test data", "Faker", "demo data", "source tables" Read data_product_accelerator/AGENTS.md
"silver layer", "DLT", "expectations", "data quality", "DQX" Read data_product_accelerator/AGENTS.md
"gold layer", "dimensional model", "ERD", "YAML schema", "merge scripts" Read data_product_accelerator/AGENTS.md
"Genie Space", "semantic layer", "metric view", "TVF" Read data_product_accelerator/AGENTS.md
"monitoring", "dashboard", "alert", "observability", "anomaly detection" Read data_product_accelerator/AGENTS.md
"MLflow", "ML model", "training", "inference", "ML pipeline" Read data_product_accelerator/AGENTS.md
"GenAI agent", "ResponsesAgent", "AI agent", "evaluation" Read data_product_accelerator/AGENTS.md
"schema CSV", "bootstrap", "new project", "build data platform" Read data_product_accelerator/AGENTS.md
"project plan", "architecture plan", "planning" Read data_product_accelerator/AGENTS.md
"Asset Bundle", "DAB", "deploy pipeline", "job YAML" Read data_product_accelerator/AGENTS.md
"naming", "tagging", "PII", "PK/FK", "constraints", "table properties" Read data_product_accelerator/AGENTS.md

Databricks AppKit Workshop — apps_lakebase/

The app is scaffolded at runtime via databricks apps init (not pre-built). The skills guide the full lifecycle.

Keywords Read This
"scaffold", "create app", "new app", "AppKit", "init app", "bootstrap app" apps_lakebase/skills/01-appkit-scaffold/SKILL.md
"build app", "implement UI", "frontend", "backend", "PRD", "components" apps_lakebase/skills/02-appkit-build/SKILL.md
"deploy app", "app.yaml", "Databricks Apps", "validate app" apps_lakebase/skills/03-appkit-deploy/SKILL.md
"Lakebase", "PostgreSQL", "database tables", "wiring", "persistence" apps_lakebase/skills/00-appkit-navigator/SKILL.md
"wire lakebase", "DDL", "CRUD API", "database schema design", "useLakebaseData", "mock fallback" apps_lakebase/skills/05-appkit-lakebase-wiring/SKILL.md
"wire agent", "agent endpoint", "serving plugin", "model serving", "useServingStream" apps_lakebase/skills/06-appkit-serving-wiring/SKILL.md
"agent app proxy", "two apps", "separate Agent App", "OBO proxy", "x-forwarded-access-token" apps_lakebase/skills/06d-appkit-agent-app-proxy/SKILL.md
"chat history", "persistent chat", "conversation sidebar", "save messages" apps_lakebase/skills/07-appkit-chat-history/SKILL.md
"feedback", "thumbs up", "thumbs down", "MLflow assessment", "rate response" apps_lakebase/skills/08-appkit-feedback/SKILL.md
"add plugin", "analytics plugin", "genie plugin", "files plugin" apps_lakebase/skills/04-appkit-plugin-add/SKILL.md
"workshop guide", "full lifecycle", "phases", "Instructions" apps_lakebase/Instructions.md

GenAI Agent Course — genai-agents/

The standalone GenAI course is the current entry point for building production GenAI agents on Databricks. Use it for Track A custom Agent Apps, MLflow GenAI foundation, evaluation, deployment, monitoring, and the AppKit 2-Apps prompt guide.

Keywords Read This
"GenAI agent course", "agent course", "Track A", "custom agent app", "canonical 2-Apps" genai-agents/00-course-orchestrator/SKILL.md
"prompt guide", "copy paste prompts", "AppKit agent app", "OBO proxy walkthrough" genai-agents/PROMPT-GUIDE.md
"agent tracing", "MLflow GenAI", "evaluation datasets", "scorers", "prompt registry", "production monitoring" genai-agents/00-course-orchestrator/SKILL.md

Agentic Framework — agentic-framework/

Keywords Read This
"multi-agent", "Foundation Model", "tool-calling loop", "agent loop" agentic-framework/skills/foundation-model-agent-loop/SKILL.md
"PRD", "product requirements", "analyze PRD" agentic-framework/agents/prd-analyzer.md
"scaffold skill", "create skill", "SKILL.md template" agentic-framework/agents/skill-scaffolder.md
"build tool", "Python tool", "agent tool" agentic-framework/agents/tool-builder.md
"test agent", "agent behavior test" agentic-framework/agents/agent-tester.md
"agent UI", "wire agent to UI", "frontend agent" agentic-framework/agents/agent-ui-wiring-prompt.md
"multi-agent orchestrator", "orchestrator build" agentic-framework/agents/multi-agent-build-prompt.md
"deploy agent", "agent deployment" agentic-framework/agents/databricks-deployer.md

Data Product Accelerator — 9-Stage Pipeline (Quick Reference)

For data product, medallion architecture, or Lakehouse tasks, the Data Product Accelerator provides a Design-First Pipeline with 77 agent skills. Read data_product_accelerator/AGENTS.md for the full routing table.

data_product_accelerator/context/*.csv
  → (1) Gold Design    — dimensional model, ERDs, YAML schemas
  → (2) Bronze         — source tables + test data (Faker)
  → (3) Silver         — DLT pipelines + data quality
  → (4) Gold Impl      — tables, merges, constraints
  → (5) Planning       — phase plans + manifest contracts
  → (6) Semantic Layer — Metric Views, TVFs, Genie Spaces
  → (7) Observability  — monitors, dashboards, alerts
  → (8) ML             — experiments, training, inference
  → (9) GenAI Agents   — agents, evaluation, deployment
Stage Keywords Skill Entry Point
1 "new project", "schema CSV", "Gold design", "ERD" data_product_accelerator/skills/gold/00-gold-layer-design/SKILL.md
2 "Bronze", "test data", "Faker" data_product_accelerator/skills/bronze/00-bronze-layer-setup/SKILL.md
3 "Silver", "DLT", "expectations" data_product_accelerator/skills/silver/00-silver-layer-setup/SKILL.md
4 "Gold tables", "merge scripts" data_product_accelerator/skills/gold/01-gold-layer-setup/SKILL.md
5 "project plan", "architecture plan" data_product_accelerator/skills/planning/00-project-planning/SKILL.md
6 "metric view", "TVF", "Genie Space" data_product_accelerator/skills/semantic-layer/00-semantic-layer-setup/SKILL.md
7 "monitoring", "dashboard", "alert" data_product_accelerator/skills/monitoring/00-observability-setup/SKILL.md
8 "MLflow", "ML model", "training" data_product_accelerator/skills/ml/00-ml-pipeline-setup/SKILL.md
9 "GenAI agent", "ResponsesAgent" genai-agents/00-course-orchestrator/SKILL.md

New project? Start at stage 1: place your schema CSV in data_product_accelerator/context/, then read data_product_accelerator/skills/gold/00-gold-layer-design/SKILL.md.


Key Rules

  1. Route first, act second. Match keywords in the tables above, then read the linked file before writing any code.
  2. Read the component-level AGENTS.md or README.md FIRST — each component owns its own detailed routing.
  3. Generated artifacts go at the repo root — never inside data_product_accelerator/, apps_lakebase/, or agentic-framework/.
  4. Framework content is read-only — skills, manifests, and server code are inputs; do not modify them unless explicitly asked.
  5. For data product tasks, always defer to data_product_accelerator/AGENTS.md — it contains the full 55-skill routing table.

IDE Compatibility

This framework is built on the open Agent Skills (SKILL.md) format and works with any AI coding assistant that can read files.

IDE / Agent How It Discovers This File File Reference Syntax
Cursor Auto-loads AGENTS.md @path/to/file
Claude Code Reads AGENTS.md (or CLAUDE.md) at repo root Reference files by path
VS Code / Copilot Reads AGENTS.md or .github/copilot-instructions.md #file:path/to/file
Windsurf Reads AGENTS.md or .windsurfrules at repo root @path/to/file
Codex Reads AGENTS.md at repo root Reference files by path
Other Point the agent to this file manually Paste file contents or path

Databricks Agent Skills are installed project-level into .agents/skills/ (gitignored) for all IDEs:

git clone --depth 1 https://github.com/databricks/databricks-agent-skills .agents/skills/databricks-skills

This follows the agentskills.io cross-agent standard and is discovered by Cursor, VS Code, Windsurf, Claude Code, and any compatible agent. See apps_lakebase/skills/01-appkit-scaffold/SKILL.md Step 1 for details.


Prompting Patterns (All IDEs)

Data Product Accelerator

I have a customer schema at @data_product_accelerator/context/booking_app_schema.csv.
Please design the Gold layer using @data_product_accelerator/skills/gold/00-gold-layer-design/SKILL.md

Databricks AppKit App

Scaffold a new Databricks AppKit app with analytics and Lakebase plugins.
Read @apps_lakebase/skills/01-appkit-scaffold/SKILL.md

GenAI Agent Course

Start the GenAI agent course using @genai-agents/00-course-orchestrator/SKILL.md.
Use @genai-agents/PROMPT-GUIDE.md when you want the canonical Track A + AppKit 2-Apps walkthrough.

Agentic Framework

Build a multi-agent orchestrator using Databricks Foundation Models.
Read @agentic-framework/skills/foundation-model-agent-loop/SKILL.md for the tool-calling pattern.

If your IDE doesn't support @ references, paste the file path or ask the agent to read it:

Read the file data_product_accelerator/AGENTS.md and follow its routing instructions.