Build a lightweight task management app using Composer, Flask, and SQLite. Track tasks—create, assign, update status—without extra overhead.Documentation Index
Fetch the complete documentation index at: https://notes.kodekloud.com/llms.txt
Use this file to discover all available pages before exploring further.
Table of Contents
- Clone the Repository
- Define the Project in Composer
- Composer Output Overview
- Review the Database Schema
- Quick Project Structure
- Key Application Code (
app.py) - Install & Run
- Inspect with DB Browser
- Use the Task Manager
- Conclusion & References
1. Cloning the Repository
Start by cloning a new, empty GitHub repo into your workspace.
projects/KodeKloud/task-manager) and open it:

2. Defining the Project with Composer
Open Composer (⌘L / Ctrl+L). Craft a precise prompt so the LLM generates only what you need: Flask + SQLite, raw SQL, no JS frameworks.
Prompt Highlights
- Objective: Proof-of-concept task manager using Flask & SQLite3.
- Features:
- Username/password authentication (Werkzeug hashing).
- Task CRUD: create, assign, update status (
Not started,In progress,Complete,Blocked,Closed). - List tasks with filters by status or assignee.
- Jinja2 HTML/CSS templates (no React/Angular).
- Requirements:
- Python 3.13+ & virtual environment
- Single or minimal tables (no migrations)
- Self-hosted development setup
Use strong hashing algorithms (e.g.,
werkzeug.security) and rotate your SECRET_KEY in production.3. Prompt Details and Composer Output
Select your model (we used Cloud 3.7 Sonnet). Composer generates:schema.sqlapp.pytemplates/(base.html,login.html,register.html,dashboard.html)static/style.cssrequirements.txt




4. Reviewing the Database Schema
Open schema.sql to inspectusers and tasks tables:
5. Quick Project Structure
6. Key Application Code (app.py)
7. Installing & Running
8. Inspecting with DB Browser
After initialization, open the SQLite file in DB Browser for SQLite to verify your tables.
9. Using the Task Manager
Browse tohttp://127.0.0.1:5000/ to see the login screen:

- Register a new user
- Create & assign tasks

- View, update, filter, or delete tasks:

10. Conclusion & References
Key Takeaways- Prompt specificity drives accurate LLM output.
- Model quality (e.g., Cloud 3.7 Sonnet) matters for code generation.