grug-brained-dev
Use when reviewing PRs, simplifying over-engineered code, judging architecture, renaming unclear concepts, or breaking an agent/human out of a too-clean path: code too fancy, too abstract, too clever, too many files, too many helpers, too many layers, too well-factored but painful to change. Embody Grug brain: complexity very bad, small words, no consultant speak, no hard pivot to opposite dogma.
86
83%
Does it follow best practices?
Impact
89%
1.21xAverage score across 3 eval scenarios
Passed
No known issues
Evaluation results
100%
29%PR Review: User Registration Module
Grug review of over-engineered registration module
Grug voice maintained
0%
100%
Correct output sections
0%
100%
No preamble
0%
100%
Meat identified
100%
100%
Fake helpers flagged
100%
100%
Bad names called out
100%
100%
Fake crystal identified
100%
100%
Single-use interfaces flagged
100%
100%
Concrete judgments
87%
100%
No balanced both-sides
100%
100%
Next bonk is small
100%
100%
Real thing kept
100%
100%
69%
-4%Notification Dispatcher: Cleanup Pass
Grug refactoring plan with small bonks
Grug voice used
0%
0%
Grug section format
0%
0%
No preamble
33%
50%
Steps are small bonks
90%
90%
Code-still-works check per step
100%
100%
No full rewrite recommended
100%
100%
Named booleans recommended
100%
40%
Fake helpers flagged
100%
100%
Duplicate eligibility logic noted
100%
100%
Meat identified
50%
50%
Concrete verdicts
75%
87%
No hard pivot
100%
100%
99%
22%Architecture Proposal Review
Grug review of microservices migration proposal
Grug voice throughout
0%
100%
Grug section headings
0%
100%
No preamble
33%
83%
Meat identified
100%
100%
Scale mismatch flagged
100%
100%
Microservices warning given
100%
100%
Fancy words translated
90%
100%
Abstraction demon-test applied
75%
100%
No consultant balance ritual
100%
100%
No 'it depends' fog
100%
100%
Concrete next bonk
100%
100%
No hard pivot to opposite
100%
100%
- Repository
- joshuadavidthomas/agent-skills
- Commit
40067f1
- Evaluated
- Agent
- Claude Code
- Model
- Claude Sonnet 4.6
Table of Contents
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.