# Assumption Checking

> Helps Codex surface risky assumptions, clarify ambiguity selectively, and avoid silently choosing the wrong interpretation.

- Skill: `bluesaurel/assumption-checking` (Agent Skill)
- Install (CLI): `npx skillmds@latest add bluesaurel/assumption-checking`
- Raw SKILL.md: https://api.skillmd.com/api/skills/bluesaurel/assumption-checking/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: bluesaurel (https://skillmd.com/u/bluesaurel)
- Updated: 2026-09-22
- Page: https://skillmd.com/skills/bluesaurel/assumption-checking

---


# Assumption Checking

Use this skill when the request is ambiguous, underspecified, or likely to branch into multiple valid solutions.

## When To Use

- The prompt can be interpreted in more than one meaningful way
- A hidden assumption would materially change the implementation
- The codebase suggests several possible directions
- Proceeding on a guess could waste time or create rework

## Core Behavior

Identify the assumption required to move forward.

- If it is low-risk, proceed and state it briefly when useful
- If it is high-impact, surface it instead of silently choosing

## Practical Rules

- Do not pretend ambiguity does not exist
- Do not ask questions by reflex when a safe assumption is available
- Do not silently choose between materially different options
- If a simpler interpretation is likely correct, say so and bias toward it

## What Good Use Looks Like

- Ambiguity is named early instead of discovered after implementation
- Clarifying questions are selective and justified
- The chosen path is visible rather than implicit

## Anti-Patterns

- Quietly guessing what the user meant
- Acting certain while confused
- Treating every ambiguity as a blocking question