105 lines
2.1 KiB
Markdown
105 lines
2.1 KiB
Markdown
# Cavepedia
|
|
|
|
A caving assistant built with [PydanticAI](https://ai.pydantic.dev/), [CopilotKit](https://copilotkit.ai), and Gemini.
|
|
|
|
## Prerequisites
|
|
|
|
- Google API Key (for Gemini)
|
|
- Auth0 account with application configured
|
|
- Python 3.13+
|
|
- uv
|
|
- Node.js 24+
|
|
- npm
|
|
|
|
## Environment Variables
|
|
|
|
### Web App (.env)
|
|
|
|
```
|
|
AUTH0_SECRET=<random-secret>
|
|
AUTH0_DOMAIN=<your-auth0-domain>
|
|
AUTH0_CLIENT_ID=<your-client-id>
|
|
AUTH0_CLIENT_SECRET=<your-client-secret>
|
|
APP_BASE_URL=https://your-domain.com
|
|
AGENT_URL=http://localhost:8000/
|
|
```
|
|
|
|
### Agent (agent/.env)
|
|
|
|
```
|
|
GOOGLE_API_KEY=<your-google-api-key>
|
|
```
|
|
|
|
## Development
|
|
|
|
1. Install dependencies:
|
|
```bash
|
|
npm install
|
|
```
|
|
|
|
2. Start the development server:
|
|
```bash
|
|
npm run dev
|
|
```
|
|
|
|
This starts both the UI and agent servers concurrently.
|
|
|
|
## Production
|
|
|
|
### Docker Build
|
|
|
|
**Web (Next.js):**
|
|
```bash
|
|
docker build -t cavepedia-web .
|
|
docker run -p 3000:3000 \
|
|
-e AUTH0_SECRET=<secret> \
|
|
-e AUTH0_DOMAIN=<domain> \
|
|
-e AUTH0_CLIENT_ID=<client-id> \
|
|
-e AUTH0_CLIENT_SECRET=<client-secret> \
|
|
-e APP_BASE_URL=https://your-domain.com \
|
|
-e AGENT_URL=http://agent:8000/ \
|
|
cavepedia-web
|
|
```
|
|
|
|
**Agent (PydanticAI):**
|
|
```bash
|
|
cd agent
|
|
docker build -t cavepedia-agent .
|
|
docker run -p 8000:8000 \
|
|
-e GOOGLE_API_KEY=<api-key> \
|
|
cavepedia-agent
|
|
```
|
|
|
|
### Without Docker
|
|
|
|
```bash
|
|
npm run build
|
|
npm run start:all
|
|
```
|
|
|
|
## Available Scripts
|
|
|
|
- `dev` - Starts both UI and agent servers in development mode
|
|
- `dev:ui` - Starts only the Next.js UI server
|
|
- `dev:agent` - Starts only the PydanticAI agent server
|
|
- `build` - Builds the Next.js application for production
|
|
- `start` - Starts the production Next.js server
|
|
- `start:agent` - Starts the production agent server
|
|
- `start:all` - Starts both servers in production mode
|
|
- `lint` - Runs ESLint
|
|
|
|
## Troubleshooting
|
|
|
|
### Agent Connection Issues
|
|
If you see connection errors:
|
|
1. Ensure the agent is running on port 8000
|
|
2. Check that GOOGLE_API_KEY is set correctly
|
|
3. Verify AGENT_URL has a trailing slash
|
|
|
|
### Python Dependencies
|
|
```bash
|
|
cd agent
|
|
uv sync
|
|
uv run uvicorn src.main:app --host 127.0.0.1 --port 8000
|
|
```
|