CoolFace
Apppublic

ujwal55/Rule-Based_Task_Planner

sourceHugging Faceupdated 1y agoView on Hugging Face
0likes
App README

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

๐Ÿ“‹ Rule-Based Task Planner

A simple rule-based AI agent that helps users break down high-level tasks into actionable substeps. Built with Python and Gradio, this project demonstrates a basic agentic AI workflow using hard-coded logic โ€” no machine learning involved.


๐Ÿš€ Demo

๐Ÿ‘‰ Launch the app on Hugging Face Spaces


๐Ÿง  What It Does

Enter a task like:

  • โ€”Plan a study session
  • โ€”Plan a workout
  • โ€”Plan a trip
  • โ€”Plan a presentation

And the app will return a predefined list of sub-tasks.

Example:

Input: Plan a study session Output: 1. Choose a topic to study 2. Gather necessary materials (books, notes, etc.) 3. Allocate a time slot for the session 4. Set specific goals 5. Review notes and summarize key points

๐Ÿ›  Tech Stack

  • โ€”Python
  • โ€”Gradio โ€“ For building a lightweight UI
  • โ€”Rule Engine โ€“ Dictionary-based mapping of tasks to steps

๐Ÿงฉ Agentic AI Concept

Although there's no ML model here, this project mimics agentic behavior using a hard-coded rule-based planner:

  • โ€”Maps user intent to structured outputs
  • โ€”Provides a decision-like structure via a rule engine

This is ideal for beginners looking to build agentic systems without needing a large language model.


๐Ÿ“ File Structure

app.py # Main app with Gradio interface taskplannings.py # It consist ruleengine description README.md requirements.txt # gradio


๐Ÿงช Run Locally

  1. 1.Clone the repo:
bash
git clone https://huggingface.co/spaces/ujwal55/Rule-Based_Task_Planner
cd Rule-Based_Task_Planner
  1. 1.Install dependencies: pip install -r requirements.txt
  1. 1.Run the app: python app.py

๐Ÿ’ก Ideas to Extend

  • โ€”Add a dropdown to choose common tasks
  • โ€”Allow user-defined tasks with a fallback plan
  • โ€”Use a small local LLM for few-shot task breakdowns

๐Ÿ‘จโ€๐Ÿ’ป Built by @ujwal55

Let me know if you want a more advanced version with a chatbot-style interface or LLM integration later.