Preprompted Data is a repository with prompts ready-to-use on demand.

• 📄 all prompts

• 😍 share your prompt

• 🤠 use with leicht (recommended)

Use With Leicht

For more information, check out the Wiki.

from leicht.prompts import get_prompt

get_prompt("basic")
# => "You're ChatGPT, …"

Use via API legacy

You can use the preprompt.py module, it's a portable version of the API from ramptix/leicht:

from preprompt import get_prompt

get_prompt("basic")
# "You're a helpful assistant…"

🚢 Source Code of preprompt.py

preprompted-data/preprompt.py

Lines 1 to 127 in 220d025

	"""Simple preprompt implementation.
	https://github.com/ramptix/preprompted-data
	"""
	import gzip
	import os
	import httpx # pip install httpx
	def make_directory() -> None:
	os.makedirs(".preprompt/", exist_ok=True)
	if not os.path.exists(".preprompt/.gitignore"):
	# Add .gitignore so the cache doesn't go to the
	# user's git repository. We're responsible!
	# Damn!
	with open(".preprompt/.gitignore", "wb") as f:
	f.write(b"*") # ignore all contents of this directory
	def fetch_prompt(name: str) -> bytes:
	client = httpx.Client()
	if name.startswith("community"):
	r = client.get(
	"https://github-discussions-api.vercel.app/body",
	params={
	"url": (
	"https://github.com/ramptix/preprompted-data/discussions/%s"
	% name.split("/")[1]
	)
	},
	timeout=None,
	)
	r.raise_for_status()
	return r.json()["body"].encode("utf8")
	r = client.get(
	f"https://raw.githubusercontent.com/ramptix/preprompted-data/main/src/{name}.md"
	)
	r.raise_for_status()
	return r.content.strip()
	def save_prompt(path_name: str, data: bytes):
	if "/" in path_name:
	os.makedirs("/".join(path_name.split("/")[:-1]), exist_ok=True)
	with gzip.open(path_name, "wb") as file:
	file.write(data)
	def read_prompt(path_name: str):
	with gzip.open(path_name, "rb") as file:
	return file.read().decode("utf-8")
	def get_cached_prompt_or_fetch(name: str, no_cache: bool = False) -> str:
	if no_cache:
	return fetch_prompt(name).decode("utf-8")
	make_directory()
	path_name = ".preprompt/%s.prompt" % name
	if not os.path.exists(path_name):
	prompt_b: bytes = fetch_prompt(name)
	save_prompt(path_name, prompt_b)
	return prompt_b.decode("utf-8")
	return read_prompt(path_name)
	def get_prompt(name: str, *, no_cache: bool = False, **kwargs: str) -> str:
	"""Get a prompt from preprompted-data.
	Args:
	name (str): Name of the prompt.
	no_cache (bool): Do not fetch and save to cache.
	**kwargs: Kwargs to fill the prompt if needed.
	Returns:
	str: The prompt.
	Raises:
	httpx.HTTPStatusError: If fetching failed, this will be raised.
	"""
	p = get_cached_prompt_or_fetch(name, no_cache=no_cache)
	if kwargs:
	for k, v in kwargs.items():
	p = p.replace("{%s}" % k, v)
	return p
	def clear_cache():
	"""Clears all prompts in ``.preprompt/*``."""
	import shutil # noqa: F401
	shutil.rmtree(".preprompt", ignore_errors=True)
	def update_all(_dir: str = ".preprompt"):
	"""Update all prompts."""
	make_directory()
	path = lambda p: p.replace("\\", "/") # noqa: E731
	for file in os.listdir(_dir):
	if file.endswith(".prompt"):
	path_name = path(os.path.join(_dir, file))
	name = path_name[11:-7] # len(".preprompt/") = 11, len(".prompt") = 7
	try:
	save_prompt(path_name, fetch_prompt(name))
	except httpx.HTTPStatusError as err:
	if err.response.status_code == 404:
	print(
	f"\x1b[1;31m[404] Prompt {name!r} is not available, skipping.\x1b[0m"
	)
	os.remove(path_name)
	elif os.path.isdir(os.path.join(_dir, file)):
	update_all(".preprompt/%s" % file)