What happens when a research phase changes?

The skill uses a PostPhaseTransition hook to automatically update or relax rules as the study progresses, such as allowing cross-case comparison only after initial coding is complete.

Where are the generated rules stored?

Rules are generated as Markdown files in the .claude/rules/ directory, ensuring they are automatically discovered and followed by the Claude Code environment.

How does saturation tracking work?

It monitors code generation, coverage, and redundancy across documents to calculate a saturation level, providing recommendations on whether more data analysis is required.

What are friction levels in this skill?

Friction levels (silent, nudge, challenge, hard_stop) determine how strictly Claude enforces methodological boundaries, ranging from simple logging to blocking actions that violate research integrity.

Can I use this for multi-researcher teams?

Yes, the skill includes tools for managing team members, logging attribution for every analytical decision, and conducting intercoder reliability sessions.

Methodological Research Rules

Name: Methodological Research Rules
Author: linxule

bylinxule

•

データサイエンスとML

Enforces qualitative research integrity through automated rule generation, saturation tracking, and workspace branching for Claude Code.

This skill acts as a methodological guardian for AI-assisted qualitative research, ensuring epistemic rigor by automatically generating and managing Claude Code rules based on your specific study design. It prevents cross-case contamination via isolation rules, monitors theoretical saturation, manages interpretive branching for alternative analyses, and tracks researcher attribution. By implementing graduated friction levels—ranging from gentle nudges to hard stops—it maintains the essential boundary between data and theory, providing a robust infrastructure for complex academic and professional qualitative studies within the Claude Code environment.

主な機能

01Graduated friction system (silent to hard stop) to enforce methodological boundaries.

022 GitHub stars

03Visual CLI dashboards for real-time monitoring of research status, rules, and analytical lineage.

04Multi-dimensional saturation tracking (generation, coverage, refinement, and redundancy).

05Automated rule generation for case, wave, and stream isolation based on project config.

06Interpretive workspace branching for exploring alternative analytical frameworks and negative cases.

ユースケース

01Monitoring theoretical saturation in Grounded Theory workflows to determine when data collection is sufficient.

02Conducting comparative case studies where cross-case contamination must be strictly prevented during early coding.

03Managing multi-researcher collaboration with intercoder reliability sessions and detailed attribution logs.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add linxule/interpretive-orchestration methodological-rules

For use in Claude.ai and ChatGPT

Download Skill