Now liveThe Skillselion MCP - thousands of ranked skills, loaded into your agent mid-task. No install.Get it →
Kuu avatar

Dig

  • 210 repo stars
  • Updated July 6, 2026
  • fumiya-kume/claude-code

Run a deep exploratory interview to surface unknowns and hidden assumptions and strengthen a plan before building.

About

Dig is a Claude Code skill that runs a deep exploratory interview to discover unknowns and challenge assumptions in a plan. A developer uses it before committing to an approach, to find blind spots and tighten the scope and requirements of the work.

  • Exploratory interview
  • Surfaces unknowns & assumptions
  • Strengthens plans

Dig by the numbers

  • Data as of Jul 25, 2026 (Skillselion catalog sync)
/plugin marketplace add fumiya-kume/claude-code
/plugin install dig@kuu-marketplace

Add your badge

Show developers this skill is listed on Skillselion. Paste this into your README.

Listed on Skillselion
repo stars210
Last updatedJuly 6, 2026
Repositoryfumiya-kume/claude-code

What it does

Run a deep exploratory interview to surface unknowns and hidden assumptions and strengthen a plan before building.

README.md

dig

Deep exploratory interview to discover unknowns and strengthen plans using the AskUserQuestion tool.

Overview

This plugin conducts deep, iterative interviews to uncover hidden assumptions, undiscovered risks, and unconsidered decisions in your plans. Unlike simple clarification, dig probes beneath the surface -- challenging premises and finding what you haven't thought of yet.

Installation

/plugin marketplace add fumiya-kume/claude-code
/plugin install dig@fumiya-kume/claude-code

Usage

/dig

Features

  • Assumption Mapping - Identifies and categorizes implicit assumptions by risk level before questioning
  • Depth-first investigation - Goes 2+ levels deep on each topic before moving to the next
  • Structured questioning - Generates 2-3 focused questions per round with concrete options including pros/cons
  • Iterative deepening - Discovers new questions from previous answers, creating a deepening spiral
  • Completeness evaluation - Uses a structured checklist to determine when investigation is thorough
  • Plan integration - Writes all decisions directly to the plan file

How it works

  1. Context Gathering - Read plan files, CLAUDE.md, and related documentation
  2. Assumption Mapping - Identify implicit assumptions and rank by risk
  3. Deep Investigation - Iterative rounds of structured questioning via AskUserQuestion
  4. Apply & Integrate - Document discoveries and update the plan file
  5. Completeness Evaluation - Check coverage and loop back if gaps remain

Allowed Tools

This command uses the following tools:

  • Write
  • Edit
  • Read
  • Grep
  • Glob
  • TodoRead
  • TodoWrite
  • AskUserQuestion

Plugin Structure

dig/
├── .claude-plugin/
│   └── plugin.json
├── commands/
│   └── dig.md
└── README.md

License

GPL-3.0

Related skills

This week in AI coding

Five minutes, every Monday - the tools, releases and tactics for developers.

unsubscribe anytime.