Researcher ORCID Identifier

0009-0000-7319-2056

Graduation Year

2026

Date of Submission

4-2026

Document Type

Campus Only Senior Thesis

Degree Name

Bachelor of Arts

Department

Mathematical Sciences

Reader 1

Mark Huber

Terms of Use & License Information

Terms of Use for work posted in Scholarship@Claremont.

Rights Information

© 2026 Kevin S Marin

Abstract

This project launches a privacy-focused command-line tool made for developers who primarily operate within  the terminal. MetricPath delivers a research-based daily score and a concise set of evidence-supported nudges, all accessible directly within the terminal, allowing users to use the prompt without leaving the development shell. A check-in collects seven daily-varying inputs: hours of sleep, cups of water, minutes of deliberate exercise, current stress, a rotating UCLA-3 loneliness item, hours spent in the terminal, and hours in agent sessions. These inputs are mapped to six health domains  using a weighted geometric mean, and adjusted by a  penalty for smoking and alcohol consumption. The output is a Day Score ranging from 0 to 1000, organized into a five-tier ladder. The CLI supports three large language model (LLM) providers: Anthropic, OpenAI, and a fully local Ollama path. By default, it functions without an API key, selects Ollama, and does not store data remotely. Data persistence is opt-in, and a single reset command erases the local history file following a typed-DELETE confirmation. This paper details the CLI’s architecture, scoring ,  check-in process, privacy model, and the design decisions underlying its development along with a user experience walk-through.

Download

This thesis is restricted to the Claremont Colleges current faculty, students, and staff.

Share

COinS