jordanfowler/messari-metrics

messari-metrics

Key takeaways

1. Used Goa (https://github.com/goadesign/goa) to generate design-based service, see: design/design.go
   • Took some extra time on this as I've been wanting to learn Goa
   • See gen/http/openapi* for autogenerated specs
2. Basic Messari API client with two endpoints, see: messari/client.go
   • https://data.messari.io/api/v2/assets used for aggregate caching
   • https://data.messari.io/api/v1/assets/{slug}/metrics used for individual metric lookup
3. Implemented Metrics Service API, see: endpoints.go
   • /aggregate for aggregate asset metrics
   • /asset/{slug} for individual asset metrics
4. Implemented an AssetCache for reduced aggregate request times, see: asset_cache.go
   • I would move this to a Redis-backed cache in a real world scenario
5. Wrote some tests, but would write a lot more in the real world

Setup

1. Check out project: git clone git@github.com:jordanfowler/messari-metrics.git
2. From project root: go mod download
3. Export your Messari API Key: export MESSARI_API_KEY=...
4. From project root: go run .

API Requests

1. Single asset metrics endpoint: localhost:8000/asset/{slug}
   • Example: curl localhost:8000/asset/btc
2. Aggregate asset metrics (query params optional): curl localhost:8000/aggregate[?tags=...|sector=...]
   • Mkt Cap >= 100M: curl localhost:8000/aggregate
   • Tag example: curl localhost:8000/aggregate?tag=NFT
   • Sector example: curl localhost:8000/aggregate?sector=Scaling

Caveats

• Aggregate endpoint will take a few seconds to have a fully warmed cache
• Until removing Messari API rate limiting, aggregate endpoint may encounter errors with sub-requests