AmpleAI Plugin is an Amplenote plugin that adds OpenAI & Ollama interactivity with Amplenote.

1. Clone this repo. git clone git@github.com:alloy-org/openai-plugin.git
2. Install node and npm if you haven't already.
3. Run npm install to install the packages.
4. Copy .env.example to .env and fill in the environment variable for your OpenAI key

To run a specific test file, use: NODE_OPTIONS=--experimental-vm-modules npm test -- test/plugin.test.js

For Cursor & LLMs to invoke a test file, the following needs to run OUTSIDE SANDBOX MODE, otherwise the tests will fail with "Error: Cannot find module '@jest/test-sequencer'"

cd /Users/bill/src/ai-plugin && NODE_OPTIONS=--experimental-vm-modules npm test -- test/path-to-file.test.js

And to invoke a test file while only running the test whose name includes "llama":

cd /Users/bill/src/ai-plugin && NODE_OPTIONS=--experimental-vm-modules npm test -- test/path-to-file.test.js  -t "llama"

Or to run allll the tests, NODE_OPTIONS=--experimental-vm-modules npm test

npm test -- -t "should allow appOption freeform Q&A" test/plugin.test.js

If you don't have Ollama running locally, you can skip the local LLM tests (which test Mistral and other Ollama models) by setting the LOCAL_MODELS environment variable to suspended:

LOCAL_MODELS=suspended NODE_OPTIONS=--experimental-vm-modules npm test

This will prevent test failures from local model tests when Ollama is not running.

https://public.amplenote.com/F4rghypGZSXEjjFLiXQTxxcR

NODE_OPTIONS=--experimental-vm-modules npm run test -- --watch

• https://esbuild.github.io/getting-started/#your-first-bundle
• https://jestjs.io/
• https://www.gitclear.com

OLLAMA_ORIGINS=https://plugins.amplenote.com ollama serve