@glowlabs-org/crm-bindings NPM

GCA CRM Backend

Elysia
- A typesafe API Client
- The types are bundled into an npm module and reused across projects to guarantee typesafe requests and responses
- NPM Publishing is handled in CI
Postgres
Drizzle ORM
Redis
- Used for caching and potentially rate limiting in future
Viem / Ethers

You can download the npm module by running

bun install @glowlabs-org/crm-bindings elysia@1.0.16 @elysiajs/eden@1.0.12

You can then go to this example to get a quickstart into consuming the API.

Swagger docs are also available here

DATABASE_URL="postgresql://postgres:<password>@<host>:<port>/<database>?sslmode=disable"
REDIS_URL=<YOUR_REDIS_URL>
MAINNET_RPC=<YOUR_MAINNET_RPC>

Install dependencies bun install

Run the dev server bun run dev 2. Get somewhat familiar with the repository structure

Index	Folder/File	Description
1	`index.ts`	Kickoff function that initiates the server.
2	`constants` folder	Contains all constants that are reused through the application
3	`routers` folder	Contains each individual microservice. Routers are grouped by group relavance. For example, all routes regarding rewards go in the `rewards` router. The prefix for a controller can be found in its respective entrypoint file. Utility functions scoped only to the router should be included in this folder.
5	`db` folder	Contains the `schema` and ORM object for the database
6.	`lib` folder	Includes all utils and helpers that are reused the project. Includes reusable utils for redis, using web3 functions, etc.
7.	`crons` folder	Includes all low level code implementation for each respective cron Job. Crons are initialized at the top level in `index.ts`.

If adding or removing columns from the database schemas, make sure to first get familiar with src/db/schema.ts. This file contains a list of rules to adhere to when making changes to database schemas.
Generally speaking, keep in mind where decimals are being manipulated and where they are not. For example, usdg rewards and glow rewards are stored with 2 decimals of precision as sql bigints due to sql size limitations. This pattern is adhered to throughout the codebase.
Also make sure that before running a database schema change, you run bun run test. Some mutations are raw sql for performance and they rely on manually coding the sql fields. An example in src/crons/update-user-rewards/update-user-rewards-for-week.ts
You should generally try to avoid writing raw SQL and use the ORM. If you are writing raw SQL for optimization purposes, make sure to add unit tests that would cause the testing pipeline to break if your mutation would break from the schema change.