Five Whys

five-whys mode

Activation: /five-whys, "activate five-whys", "five whys mode". Deactivation: /default, "deactivate", "normal mode".

Protocol

Before writing any fix, output exactly this structure:

Why 1:  → 
Why 2:  → 
Why 3: → 
Why 4: → 
Root: 
Fix: 

Rules:

• Every Why must follow logically from the previous. No leaps.
• If a Why hits "unknown", run a tool (Read/Grep/Bash) to find out before continuing.
• Stop early only if the chain genuinely bottoms out before 5 (rare). Note "bottomed at Why N".
• The Fix targets the root, not Why 1.

Override

If user says "just fix it", "skip the whys", or similar at any point: acknowledge in one line ("Skipping chain — fixing directly."), then fix. User always has override.

When to apply

• Debugging a bug
• Performance issue
• Architecture decision
• Recurring incident
• Code review of a fix ("does this address the root?")

When to skip silently

• Trivial typo / rename / formatting
• New feature with no underlying problem to diagnose
• User explicitly asked for a quick patch

Boundaries

• Code correctness, syntax, error messages: unchanged
• Don't invent causes — if a Why requires guessing, investigate first
• Safety-critical bug → still run the chain, but state the immediate mitigation in line 1 before the chain

Token note

Adds ~150-300 tokens per fix. Worth it for non-trivial bugs.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: repowise-dev
• Source: repowise-dev/modpack
• License: MIT
• Homepage: https://repowise.dev

Install and usage instructions live in the source repository linked above.