truffle-ai-backend

Getting Started

You can run the application locally using NodeJS or Docker.
For development we recommend using NodeJS, for production we recommend using Docker with Kubernetes.

Important - Environment variables

Before continuing make sure you have set up the environment variables, for Docker in a file called .env and for NodeJS in a file called .env.dev

Running the app locally with Docker (production)

We recommend you to install Docker for production.
To build all Docker images and serve them in containers, run

cd truffle-ai-backend # go to project
docker compose up

Your production ready images will now be created and served locally. You can run them in a Kubernetes cluster, see config/deployment.template.yml for configuration.

Developing & Building the app with NodeJS

To get started with the development of the application, install NodeJS.
Then run the following in a terminal to install all dependencies.

cd truffle-ai-backend # go to project
npm ci # install dependencies

To run the application in development mode, run

npm run dev:services # dev mode

To build and serve the application with NodeJS, run

npm run build:services # builds the app
npm run serve # serve the app

Packages

The project consists of three microservices

• graphql_gateway a gateway composing multiple GraphQL schemas into one, authentication check with jwt cookie
• graphql_server a Graphql server
• repo_job a CronJob running at 2am to fetch data from APIs

graphql_gateway

The gateway service composes schemas into one superschema and redirects the queries to the corresponding service. This is currently a postgraphile Supabase and the graphql_server service.

graphql_server

The graphql server service sets up a Fastify server, that runs Mercurius to host a GraphQL schema. It will be used to build our own GraphQL schemas.

npm run update-types # updates supabase graphql types

repo_job

The repo job service runs a script several times a day that notifies the graphql_server to update all bookmarked repositories.

To work with all services simultaneously you can use NPM workspaces.

List of Commands

Note: Services are served continuously and Jobs are run once.

npm ci # install all dependencies from package-lock.json

npm run prepare # installs pre-commit hook (running lint-staged on staged files)
npm run dev # start services and jobs in dev mode
npm run dev:jobs # start jobs in dev mode
npm run dev:services # start services in dev mode
npm run build # build services and jobs
npm run build:jobs # build jobs
npm run build:services # build services
npm run serve # serve the built services (excluding jobs)
npm run job # starts all jobs

npm run test # check your files format and lint them
npm run format # format all your files with Pettier
npm run lint # lint your files with ESLint

docker compose build # builds the Docker containers
docker compose up # builds and runs created Docker containers

Environment Variables

Before starting to develop and build check that you have setup your environment variables in .env.dev for NodeJS or .env for Docker-Compose.

SUPABASE_REFERENCE_ID # supabase reference id
SUPABASE_URL # supabase url
SUPABASE_GRAPHQL_URL # supabase graphql url
SUPABASE_API_KEY # supabase api key
SUPABASE_SERVICE_API_KEY # supabase service api key
GATEWAY_PORT # port number
SERVER_PORT # port number
SERVER_URL # url to graphql_server
SERVER_GRAPHQL_URL # url to graphql route on graphql-server
NODE_ENV # development|production
GITHUB_API_URL # github api url
GITHUB_API_TOKEN # github api token
OPENAI_API_KEY # openAI api key
SCRAPING_BOT_API_KEY
SCRAPING_BOT_USER_NAME
SCRAPING_BOT_API_ENDPOINT