examples/sample_prompts.md

Sample Prompts for NexaSci Agent

These prompts are designed to showcase the agent's capabilities: reasoning, tool usage, code generation, and citation.

Python Simulation Prompts

1. Physics Simulation

Design a Python simulation to model the trajectory of a projectile launched at 45 degrees with air resistance. Include visualization and explain the physics.

2. Diffusion Model

Create a Python simulation to model the diffusion of nanoparticles in a fluid. Include visualization and cite relevant literature on nanoparticle transport.

3. Chemical Reaction

Simulate a first-order chemical reaction A → B with rate constant k=0.1. Plot concentration vs time and compare with analytical solution.

Literature Search & Citation Prompts

4. Recent Advances

Summarize recent advances in room-temperature superconductivity. Search for papers from the last 2 years and cite key findings.

5. Methodology Review

Find and summarize papers on CRISPR-Cas9 gene editing techniques. Focus on delivery methods and cite at least 3 recent papers.

6. Comparative Analysis

Compare different approaches to quantum error correction. Search the literature and provide a structured comparison with citations.

Experimental Design Prompts

7. Battery Research

Design a reproducible procedure to measure battery degradation in solid-state lithium cells over 200 charge cycles. Include controls and statistical analysis.

8. Material Synthesis

Design an experiment to synthesize graphene oxide using the Hummers method. Include safety considerations and quality characterization steps.

9. Biological Assay

Design a lab protocol to measure protein concentration using the Bradford assay. Include calibration curve preparation and error analysis.

Combined Reasoning Prompts

10. Full Workflow

I want to understand how machine learning is being used in drug discovery. First, search for recent papers on this topic, then write a Python script to analyze a sample dataset of molecular properties, and finally summarize the key findings with citations.

11. Code + Literature

Create a Python simulation of a neural network learning XOR, then search for papers on neural network interpretability and cite relevant work.

12. Hypothesis Testing

Formulate a hypothesis about the relationship between temperature and enzyme activity. Design an experiment to test it, write Python code to simulate the results, and search for literature to support or refute your hypothesis.

Usage

Run any of these prompts with:

python examples/demo_agent.py --prompt "Your prompt here"

Or use the test script:

python examples/test_agent.py --raw "Your prompt here"