Ekubo API

This repository contains the code for the Ekubo API, which powers all aspects of the interface.

Running local

All you need to run this API locally is a database containing the schema defined in the EkuboProtocol/indexer repository. Follow instructions in the indexer repo to set up the database.

Once you have a database that is populated with data, run bun install, and then bun run start.

The local connection string defaults to postgresql://postgres:postgres@localhost:5432/postgres via [vars] in wrangler.toml. Override it by putting PG_CONNECTION_STRING in a .dev.vars file, which is gitignored.

Deploying your own

wrangler.toml is configured for Ekubo's own Cloudflare account, so a self-hosted deployment needs three changes under [env.prod]:

• hyperdrive.id — the c6bd83f2… binding points at Ekubo's Hyperdrive config. Create your own over your indexer database (wrangler hyperdrive create) and use that id. Hyperdrive holds the database credential, which is why none appears here.
• routes — prod-api.ekubo.org is a custom domain on Ekubo's zone. Point it at a domain you control, or drop the block and use the workers.dev URL.
• name — ekubo-api is the Worker name within an account; change it if it collides.

The Worker needs no secrets of its own: the only bindings it reads are HYPERDRIVE in production and PG_CONNECTION_STRING locally. Deployment from CI additionally uses the CF_API_TOKEN and CLOUDFLARE_ACCOUNT_ID repository secrets; the deploy job is skipped on forks, so tests still run green without them.

License

MIT — see LICENSE.