Here’s all the functions in the Pack. The Pack is very much a work in progress, so generated outputs are pretty rough. I think so far the best use is the Classify formula if you have a bunch of data. Or the Expand formula for expanding on brainstorming. Prompt for the following 3 functions:Here's why I think you should buy houseplants:
Generate generates text content, it should continue from the prompt
🫤 I tried to make GenerateParagraph generate a paragrah continuing from the prompt but results seem not quite paragraph length
🫤 I tried to make GenerateSentence generate a sentence continuing from the prompt but results seem worse than the base Generate formula because of the extra fluff I put around the prompt to specify that only a sentence should be generated.
Prompt for the following 5 formulas:
- Breathing is important
- Breathing through your nose is best
🫤 I tried to make GeneratePoints generate a set of points based on another set of points but results seem not quite related and also not quite in the right format
🫤 I tried to make GenerateStory tell a story based on a prompt but results seem not quite story like or length
Expand expands upon a some bullet points; although some points seem like incomplete thoughts
🫤 I tried to make Rephrase rephrase whatever the input was but it doesn’t really capture the same meaning
Related seems to do okay at generating related topics based on a prompt
Prompt for the following formula:
A Pack is an extension that adds new powers to your doc. Packs work by supplementing the core building blocks of Coda, so you can customize how your doc looks, works, and integrates.
With a little JavaScript, anyone can create and publish a Pack to the Gallery. All you need to do is write the code and Coda will deploy and host the Pack for you.
Packs can be created for personal use, shared with a team, or published to the world. Everything from simple utilities to rich integrations can be built with Packs.
🫤 I tried to make Summarize summarize the prompt but results seem to not quite capture the content
Classify uses input training data to then classify other inputs
Classification "training" data
Nice to see you, I'm glad you're here
is a
Tokenize tokenizes
Nice to see you, I'm glad you're here