# ai-hackathon-basic-tools

## prep

- harvest context
- make sure accounts and API keys are available
- work on prepping Video Games workshop
- work on prepping Wednesday workshop
- talk about different working styles on the team

### mk

- slack app for The Chronicle Lab
- prep for Christine and MK's stations at video games
- Claude Code
- Cursor
- VS code
- bring computer for dani's team
- github repo for each
-
-

## Basics for Everyone

- make sure you are in our bok slack channel ``ll-project-huit-ai-hackathon-2025`` . this is where we will communicate about this project!
- only leaders of teams get API keys
- we'll determine who needs github access from team to team, but let's add everyone who wants to be added to a new all-ll organization where we can create the repos for each project
- team leads should log in to the HUIT tools but fellows, undergrads and faculty don't need to
- every machine we're using needs the usual
- vs code
- git
- homebrew
- node
- optionally cursor
- some groups may be testing out a specific tool/plugin for vs code (claude code, codex) and then should install that and connect
-

## what is “Agentic AI”

### HUIT's links
- Read one explainer:
- [IBM: What Is Agentic AI?](https://www.ibm.com/think/topics/agentic-ai)
- [Stanford: Enhancing Your Understanding of Agentic AI](https://online.stanford.edu/enhancing-your-understanding-agentic-ai-practical-guide)

## Per-Team Suggestions OLD

### The Capture Lab
**Lead:** Jordan
**Focus:** Real-time agents for transcription, tagging, and vision.

**Tools:**
- OpenAI or Gemini APIs for image recognition and speech-to-text.
- Python (for quick prototyping) or Node.js.
- A live transcription tool like Whisper or AssemblyAI.
- Markdown basics for auto-formatted summaries.

**Prep:**
- Test uploading an image to an API and getting a description back.
- Practice transcribing a short video or audio clip with Whisper.

---

### The Chronicle Lab
**Lead:** Dani
**Focus:** Slack and Airtable research agents that summarize and reflect on Lab activity.

**Tools to review:**
- Airtable API and schema of the Bok Lab base.
- Slack API / app setup (for reading and posting messages).
- Python or JavaScript basics for bot building.
- OpenAI or Claude APIs for summarization or reflection.

**Prep:**
- Download and inspect the [AI Links and Ideas Bot](https://github.com/ll-color-machine/slack-links-and-ideas-bots).
- Create a `.env` file for API keys.
- Try running the bot locally to ensure setup works.
- Look at the Agent Builder, and think about: if we were going to try to reproduce the Agent Builder in Airtable, how would we do it? What fields, relationships, or tables would we need?
- [News bots:](/2Pjse49MROaTgKYVePNvZQ)
---

### The Dialogue on Display Lab
**Lead:** Madeleine
**Focus:** Conversational agents in a visual interface.

**Tools to review:**
- Next.js (front-end framework).
- OpenAI or Claude for multi-agent conversation.
- Basic HTML/CSS and simple state management (React).

**Prep:**
- Fork a simple Next.js chat example.
- Test connecting an API endpoint that returns model output to a web page.

---

### The Composition Lab
**Lead:** Christine
**Focus:** Agentic creative tools for multimodal academic essays.

**Tools to review:**
- Claude Code and Codex for code-assisted writing.
- Gemini CLI (if available) for multimodal text/image support.
- Cursor IDE (if not using VS Code).
- Web-based publishing tools (Next.js, GitHub Pages, or Glitch).

**Prep:**
- Create a new local Next.js project and deploy a “hello world” page.
- Ask Claude Code or Codex to generate sample code blocks for a text/image essay layout.