rocketride-org/rocketride-server ? reverse-engineered prompt

Build me an open source AI workflow tool that turns my editor into a place where I can design, run, debug, and deploy AI pipelines without stitching a bunch of services together by hand.

I want a visual pipeline builder in VS Code, plus a CLI for running the same flows from the terminal. Pipelines should be saved in a portable JSON or pipe style format, run locally, in Docker, or against a hosted endpoint just by changing config. Please include a fast runtime, good observability so I can see what each step is doing, and a library of reusable nodes for common AI tasks like model calls, vector search, OCR, NER, and agent style orchestration. It should work with multiple model providers and vector databases, and I should be able to drop a pipeline into a Python or TypeScript app with simple SDK examples.

Please wire up sensible env templates, example pipelines, and enough tests and docs that I can actually use it. If you need details, check the current docs online and follow the repo’s intended setup.