Mentor.skill

Distill Mentor - Academic Mentor Digital Avatar Generation Tool

Skills Overview

Distill Mentor is a Claude skill that automatically collects information about academic mentors, analyzes their research styles, and generates conversational AI mentor capabilities.

Use Cases

1. Create a Digital Avatar of an Academic Mentor

When you want to preserve your mentor’s academic experience and research style, or need to ask questions even when the mentor is unavailable, you can use Distill Mentor to automatically collect the mentor’s papers and homepage information. It performs a deep analysis of their research methodology, writing style, and academic values, ultimately generating an AI skill that can answer questions in the mentor’s style.

2. Analyze a Researcher’s Style and Academic Preferences

Researchers and students, when writing papers or planning research, can use Distill Mentor to analyze a specific scholar’s research preferences, including: the types of problems they choose (theoretical vs. practical), experimental design patterns, dataset preferences, visualization styles, writing structure, and more. This helps you learn from and emulate outstanding researchers’ ways of thinking.

3. Generate Domain Mentor Knowledge Bases in Bulk

For academic institutions or research teams, Distill Mentor can create digital skill libraries for multiple mentors or domain experts. This forms a collective wisdom system that is searchable and conversational, helping new members quickly understand differences in research styles and methodologies within the field.

Core Features

1. Automatic Collection from Multiple Sources

Using the ArXiv API, web search, and personal homepage scraping, it automatically collects the target mentor’s paper list, research descriptions, and public information. It supports precisely locating same-name scholars by institution name and can handle additional materials uploaded by users, such as CVs and publication lists.

2. Deep Paper and Style Analysis

Based on collected papers and public information, it uses structured analysis prompt templates to extract 10+ dimensions of a style profile: research interests and evolution, methodology preferences, writing style characteristics, academic values, communication patterns, teaching philosophy, and more. It supports bilingual analysis in both Chinese and English.

3. Generate a Conversational Skill with One Click

The analysis results are automatically packaged into a standard Claude skill file (SKILL.md) and a JSON configuration. The generated skill can be loaded and used directly, and it conducts conversations using the mentor’s typical modes of expression, thinking patterns, and feedback style. It supports incremental updates and iterative optimization.

FAQ

What is Distill Mentor?

Distill Mentor is a Claude skill used to “distill” an academic mentor’s knowledge and style. It can automatically collect a mentor’s papers and public information, analyze characteristics such as research style, writing approach, and academic values, and then generate an AI skill that can simulate the mentor’s way of answering questions.

How do I use Distill Mentor to generate a mentor skill?

Run /distill-mentor <mentor name> or provide institutional information with /distill-mentor <mentor name> --affiliation <institution name> to start. The system will automatically search for papers, analyze style, generate the skill files into the ~/.claude/skills/ directory, and then preview the results and request confirmation.

How accurate are the generated mentor skills?

Accuracy depends on the amount of information available. For mentors with sufficient ArXiv papers (more than 5) and public homepage information, the deep analysis feature can extract fairly detailed style characteristics. If information is insufficient, the system will use default settings. We recommend that users provide a CV, paper list, or homepage links to improve quality.