Skip to content
Claude Merch

CivicBrief

We've built a highlight reel for democracy. Unbiased, easy to digest political content for people who care about their country, but don't want to sit through hours of drawn out political drama videos.

Core trick

Uses AI APIs to locate key moments in long political videos and cut clips that link back to the original source.

Tags

Built with

claude

Description

Inspiration Most people don't tune out of politics because they don't care, they tune out because it's exhausting, polarizing, and inaccessible, and we built CivicBrief to fix that. What it does CivicBrief transforms long-form political content into short, unbiased, digestible clips that link back to the original source so users can always go deeper. How we built it We used AI APIs to automatically identify key moments in long-form video, extract meaningful clips, and surface them in a clean, scrollable interface. Challenges we ran into The hardest problem wasn't technical, it was editorial, specifically how to ensure every clip was fair, accurate, and never misleading without its original context. Accomplishments that we're proud of We built a working prototype that surfaces genuinely meaningful political moments rather than just the loudest or most dramatic ones, without telling users what to think. What we learned Building CivicBrief taught us that bias isn't always intentional, sometimes it's just an incomplete truth, and that the hardest problems in civic tech are rarely the technical ones. What's next for CivicBrief We're expanding to cover speeches, town halls, and congressional hearings, with personalized feeds based on issues users care about so any first-time voter can get up to speed in under 5 minutes.
shreyashmishra/anthropic-hackathon★ 01 forksTypeScriptCreated Mar 29, 2026 · last push Mar 30, 2026

README excerpt

# OpenFloor

OpenFloor is a civic video platform for the Governance & Collaboration track. It helps people
understand long political and public-interest recordings by turning them into timestamped summaries,
notable statements, clip-based discussions, and source-aware public threads.

The current repo includes:

- A Next.js + TypeScript + Tailwind app shell
- Built-in demo-user access for posting and participation
- Gemini-powered video processing hooks for transcript analysis and clip generation
- Drizzle ORM with a lightweight SQLite-backed local prototype
- Public browsing for videos, clips, sources, and topics
- Local-storage utilities for draft comments, UI preferences, and recently viewed content

## Product Direction

OpenFloor is designed to:

- Make civic video easier to understand
- Preserve provenance and surrounding context
- Distinguish official-source content from user submissions
- Support constructive, threaded public discussion
- Reduce decontextualized outrage bait

The detailed product blueprint lives in [docs/openfloor-blueprint.md](docs/openfloor-blueprint.md).

## Getting Started

1. Copy `.env.example` to `.env` and fill in Gemini and optional demo-user labels.
2. Install dependencies with `npm install`.
3. Push the local schema with `npm run db:push`.
4. Start the app with `npm run dev`.

Then open [http://localhost:3000](http://localhost:3000).

## Useful Scripts

```bash
npm run dev
npm run build
npm run typecheck
npm run db:push
```

## Notes

- The

Team