Ask HN: How are you managing your prompts?
How are you cataloging, maintaining, versioning your prompts? Are you relying primarily on one toolchain where they are organizing by default in the toolchain interface (but locked in the toolchain)? Or are you using some other mechanism? Are you using a mechanism that enables you to share and permission them with team members, colleagues, friends?
The intent with this inquiry is to improve my own workflow(s) and potentially those of others interested in this topic. Thank you!
(
8 comments
| 3 points
| on HN
)
These are two very different questions.
To keep prompts organized as a software startup is a completely different use-case, as we need: - a way to dynamically fill the prompts (we use Mustache) - a way to store the prompts that enable versioning (so they are on git like the rest of our code) - a way to allow non-technical users (eg. Product team) to revise a prompt, so they are stored as JSON objects.
So our prompts are basically an object that encapsulates the OpenAI-style parameters, plus additional in-house parameters such as fallback model, risk profile, etc.
[0]: https://anythingllm.com/desktop
I just prompt as I go and find that the "cost" of prompting again to get a better output is lower than the cost of having some system for cataloging, maintaining, and versioning my prompts.
I might be wrong but I'm getting good results out of LLMs.
Or a personal assistant. I have a text based workflow but explaining it to an llm takes yet another 1000 word essay. I would benefit from a workflow that lets me reuse and version prompts.
Ask HN: How do you manage your prompts in ChatGPT? https://news.ycombinator.com/item?id=41479189
I'm curious to see how people's workflows have changed.
Us, we manually catalog them in well-named Markdown files and folders and store in a git repo. I would like a more taxonomical approach.