Core import for prompt maker

Best practice to use the prompt

Basic Imports

from fastcore.all import L

source

L

 L (items=None, *rest, use_list=False, match=None)

Behaves like a list of items but can also index with list of indices or masks

source

H1

 H1 (*c:fastcore.xml.FT|str, cls:enum.Enum|str|tuple=(), **kwargs)

H1 with styling and appropriate size

	Type	Default	Details
c	fastcore.xml.FT \| str		Contents of H1 tag (often text)
cls	enum.Enum \| str \| tuple	()	Classes in addition to H1 styling
kwargs	VAR_KEYWORD
Returns	FT		H1(…, cls=‘uk-h1’)

Initializing app with live debug mode.

Basic components and temp for import

source

apply_preset

 apply_preset (preset_type)

code

source

generate_prompt_content

 generate_prompt_content (criteria, components, model='openai/gpt-4o')

source

parse_llm_response

 parse_llm_response (content)

index page

source

get

 get (preset:str='')

source

get_progress

 get_progress (session_id:str)

source

generate

 generate (request)

@rt("/generate", methods=["POST"])
async def generate(request):
    form_data = await request.form()
    criteria,components,model = form_data.get("criteria", ""),form_data.getlist("components"),form_data.get("model", "gpt-3.5-turbo")
    role_text,task_text,format_text,examples_text = form_data.get("role_text", ""),form_data.get("task_text", ""),form_data.get("format_text", ""),form_data.get("examples_text", "")
    if not criteria.strip(): return Div("Please enter some criteria first!", cls='text-red-500')
    session_id = str(uuid.uuid4())
    progress_state[session_id] = {"progress": 10, "done": False}
    form_json = json.dumps({"criteria": criteria, "components": components, "model": model, "role_text": role_text, "task_text": task_text, "format_text": format_text, "examples_text": examples_text, "session_id": session_id})
    return Div(Progress(value=10, hx_get=f"/progress/{session_id}", hx_trigger="load, every 500ms", hx_swap="outerHTML"), Div(id="final-output", hx_post=f"/generate-result", hx_trigger="load delay:100ms", hx_vals=form_json))

source

generate_result

 generate_result (request)