AI Transformation Workshop
https://github.com/coleam00/ai-transformation-workshop- Category
- Developer Tools
- Rank
- No. 1062Tools index
Previous survey · No. 1115 ·
- Pricing
- Open Source
- Type
- TOOL
- Builder
- coleam00
- GitHub
- 135 stars
- Date
About
Workshop materials for 'Principles of Agentic Engineering' by Cole Medin — the AI layer concept, the PIV loop, and 15 reusable Claude Code commands.
What it does
This is a practical teaching kit for turning coding-agent habits into a repeatable team process. It pairs documented planning, implementation, and validation practices with a working poll application. Teams can copy the agent instructions into another repository, tailor the project rules, and keep workflow improvements under version control.
Why it's ranked here
The repository makes its methodology inspectable through a runnable example, reusable workflows, and explicit validation layers. That combination gives teams something concrete to test, not just workshop theory. Its narrow Claude Code orientation and intentionally incomplete demo make it better as a reference system than a finished product.
What's good
Planning requires a validation strategy before coding starts, then implementation begins with a fresh context. Research can run in parallel while implementation stays centralized. The demo reinforces the guidance with strict TypeScript, schema validation, typed service errors, feature tests, structured logging, and a vertical feature slice that separates database work from business logic.
Tradeoffs
The poll application deliberately omits live result updates, authentication, and external services. Duplicate-vote prevention relies on cookies, so it is not a strong identity control. The workflow is shaped around Claude Code, while issue creation requires configuring an MCP connection. Human testing and code review remain explicit upper validation layers.
How to use it well
It suits engineering teams standardizing how coding agents gather context, write plans, implement changes, and validate results. Start with the poll example, adapt the rules to your stack, then version the workflow beside application code. Use it as an operating template and workshop reference, not as a production polling service or general project-management platform.
Technical notes+
The private Bun package in package.json runs a Next.js 16 and React 19 application with TypeScript 5, Biome, Bun tests, Drizzle ORM, better-sqlite3, Zod, and Tailwind CSS 4. The portable agent layer lives under .claude/, with 15 command workflows and two skills; .mcp.json wires Atlassian MCP for Jira and Confluence. The demo is under src/, while src/features/polls/ forms a vertical slice containing models, validation schemas, repository access, services, typed errors, Server Actions, cookie-based vote protection, components, and tests. src/core/api/errors.ts maps Zod and typed HTTP errors into JSON responses, and src/core/config/env.ts defaults storage to file:./local.db.
Observed
- Primary language
- TypeScript
- Packaging and install
- Private Bun package installed with bun install
- Application interface
- Next.js web application for creating, voting on, and viewing polls
- Agent interface
- 15 reusable Claude Code command workflows and two skills
- External integration
- MCP configuration for Jira and Confluence through Atlassian
- Storage
- Drizzle ORM with a local better-sqlite3 database by default
- Testing and checks
- Bun test runner, strict TypeScript, and Biome linting
Read from README.md, package.json, src/app/page.tsx, src/lib/utils.ts, src/app/layout.tsx, src/shared/index.ts, src/types/testing-library.d.ts, src/components/theme-toggle.tsx, src/components/theme-provider.tsx, src/core/api/errors.ts, src/core/config/env.ts, src/core/logging/index.ts, src/shared/utils/index.ts, src/shared/utils/dates.ts, src/core/database/index.ts.
What it can do
Teach AI layer concept implementation
Workshop materials and curriculum → Understanding of AI layer architecture principles
Guide through PIV loop methodology
Workshop exercises and examples → Practical PIV loop implementation skills
Provide 15 reusable Claude Code commands
Workshop command library → Ready-to-use Claude automation commands
Deliver agentic engineering training
Workshop curriculum and materials → Agentic engineering competency and knowledge
Enable AI transformation workflow design
Workshop principles and frameworks → Custom AI transformation strategies
Tags
Tech Stack
Comments (0)
No comments yet
Editorially curated, with community endorsements as a secondary signal. Corrections welcome.