# Beginner AI Project

## Problem and intended user

Describe one narrow task, who uses the result, and what the project does not do.

## Minimum finished version

State the smallest end-to-end deliverable in one sentence.

## Data and permission

- Source:
- License or permission:
- Fields used:
- Personal or confidential data excluded:

## Setup and reproduction

List the environment, dependencies, exact install command, and exact run command. Never include API keys.

## Baseline and method

Describe the simple baseline first, then the model, prompt, or workflow being tested.

## Evaluation

- Held-out test cases:
- Metric or rubric:
- Baseline result:
- Project result:
- Failure categories:

Attach the completed `beginner-ai-project-test-cases.csv` or an equivalent permitted test set.

## Limitations and safe fallback

Document inputs that fail, decisions the project must not make, and when it should abstain or require human review.

## Evidence

Link to runnable code, screenshots, a short demo, and representative failures. Remove secrets, personal information, customer content, and proprietary documents before publishing.

## Version history

Record the date, dataset version, model or dependency version, and the one change tested in each iteration.