Deploy Turbo pipelines from Goldsky AI
Follow @goldskyio on Twitter to stay updated on everything we ship.
Goldsky AI can now take a pipeline from idea to running. Describe what you want, and it writes the Turbo YAML, validates it, shows you the pipeline, and deploys it from the chat. You no longer have to copy the result into the editor to ship it.
Google Cloud Pub/Sub sink
Turbo pipelines can now publish to Google Cloud Pub/Sub topics, which makes it easier to feed onchain data into services already running on Google Cloud.
Local forks that work on popular chains
goldsky compose start --fork-chains now works on the chains people use most. Forking needs eth_getProof, which many public RPCs don't support, so Compose now falls back to Goldsky-hosted RPC.
The new --impersonate flag lets you run a fork as any wallet address, no private key needed. It's useful for reproducing an issue against a real wallet. See the Compose CLI reference.
- Compose CLI binaries are now available for Linux ARM64 and macOS x64
- Each Compose task run now has its own logs, so you can read one run at a time
- Fixed the next steps shown after
goldsky compose init - The app refreshes its data again when you switch back to the browser tab
- Clearer copy when you delete a Postgres sink that you host yourself
- Fixed subgraph version counts overlapping the status on the subgraphs dashboard