Skip to content
Back to skills

Assumption Tracker

ASecurity

Explicitly track, test, and validate assumptions - prevent blind spots

  • 666 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added September 2, 2026
ai-agentsrustgotestingapidatabase

Works with

  • cli
  • api

Security analysis

A100/100

Scanned September 2, 2026

npx -y skills add majiayu000/claude-skill-registry --skill assumption-tracker --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Assumption Tracker?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Assumption Tracker
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/majiayu000-assumption-tracker/badge)](https://www.skillsdirectory.com/skills/majiayu000-assumption-tracker)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: assumption-tracker
description: Explicitly track, test, and validate assumptions - prevent blind spots
category: Thinking Analysis
---

# Skill: assumption-tracker

## What I do

I surface and manage hidden assumptions. I ensure that every leap of faith in a design or plan is documented, tiered by risk, and systematically validated through evidence or testing.

## When to use me

- Before starting a new feature or architectural change
- When requirements are ambiguous or "common sense" is invoked
- During technical planning sessions to identify "we think" vs "we know"
- When evaluating third-party libraries or external API behaviours

## Core principles

1. **Surface the hidden** — If it isn't proven, it's an assumption.
2. **Tier by risk** — Focus validation on assumptions with high impact and low certainty.
3. **Validate early** — Use spikes, prototypes, or data lookups to turn assumptions into facts.
4. **Document outcomes** — Record whether an assumption was proven true or false.

## Patterns & examples

**Assumption Logging Format:**
| Assumption | Impact (H/M/L) | Certainty (H/M/L) | Validation Method | Status |
| :--- | :--- | :--- | :--- | :--- |
| "The legacy API supports concurrent writes." | High | Low | Run concurrency spike test | Pending |
| "Users prefer the sidebar over the top nav." | Medium | Medium | Review GA click maps | Validated |

**Validation Techniques:**
- **Spike:** Write a small, throwaway script to test a technical hypothesis.
- **Prototype:** Build a minimal UI to verify user interaction assumptions.
- **Data Lookup:** Query logs or databases to confirm usage patterns.

## Anti-patterns to avoid

- ❌ **"Trust me" logic** — Relying on seniority instead of evidence.
- ❌ **Validation lag** — Building a full system on unverified, high-risk assumptions.
- ❌ **Silent assumptions** — Failing to voice doubts during the planning phase.

## KB Reference

`~/vaults/baphled/3. Resources/Knowledge Base/AI Development System/Skills/Thinking-Analysis/Assumption Tracker.md`

## Related skills

- `critical-thinking` — Rigorous analysis of claims
- `epistemic-rigor` — Distinguishing belief from knowledge
- `prove-correctness` — Evidence-based validation
- `pre-action` — Deliberate thinking before execution

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.