| mode |
|---|
agent |
We are in the folder for a new project. Your job is to set up the workspace according to my usual preferences. Here are the requirements:
{
"scripts": {| #!/bin/bash | |
| # Authored by: Kagi Assistant | |
| # https://kagi.com/assistant/6995ca12-6df0-4f88-9d91-848e993a2870 | |
| # Source NVM to make it available to the script | |
| export NVM_DIR="$HOME/.nvm" | |
| [ -s "$NVM_DIR/nvm.sh" ] && \. "$NVM_DIR/nvm.sh" # This loads nvm | |
| # Check if NVM is available | |
| if ! command -v nvm &> /dev/null; then |
An AI agent should attempt to run cli commands on behalf of the user when they are needed.
Before running any command as a terminal command, the AI agent should check if there is a relevant MCP (Machine Command Protocol) command available in the project. If there is, it should use that instead of running a terminal command directly.
Check type safety using the npm run typecheck task instead of tsc directly.
The database schema can be found in schema.ts.
For timestamp and date column types use { mode: "date" } to give the column the Date type.
| { | |
| "name": "typefabric-management-app", | |
| "version": "0.1.0", | |
| "lockfileVersion": 3, | |
| "requires": true, | |
| "packages": { | |
| "": { | |
| "name": "typefabric-management-app", | |
| "version": "0.1.0", | |
| "dependencies": { |
A reusable pattern for building type-safe, validated forms in React using:
defineForm(schema): Generates a custom hook for your form, automatically wiring up Zod validation and type inference.