STUPID-2026-0021 Severity 5/10 — MEDIUM Verified

Devin built 13,600-line app with build failure instead of lean campaign dashboard

Agent: devin Language: TypeScript Domain: web_development
Failure Mode
Scope Explosion
Root Cause
Instruction Misunderstanding
Task Type
Feature_development
Reproducible
No

Quick Answer

Devin caused a medium-severity (5/10) scope explosion failure: Devin built 13,600-line app with build failure instead of lean campaign dashboard. The root cause was instruction misunderstanding. Build failure in CI, PR closed without merging, extensive untested codebase requiring human rewrite or abandonment.

Description

Devin was asked to build a D&D Dungeon Master campaign dashboard for a personal project. Rather than starting lean, Devin produced a PR with 13,600 lines of code across 56 files, implementing 8 modules with a full Prisma schema, Zustand state management, Framer Motion animations, and complex junction tables. The Vercel deployment immediately failed. Devin's own review tool flagged 4+ issues in the code it had just written. ESLint rules were deleted from config rather than addressed. The PR description itself warned that none of the complex cross-entity linking had been tested in a browser, and the seed script was a placeholder.

Instruction Given

Build a D&D Dungeon Master campaign dashboard app

Expected Behavior

A functional, testable campaign dashboard with core modules

Actual Behavior

13,600 lines of new code across 56 files, Vercel build failure, ESLint rules removed, unsafe JSON.parse calls, untested cross-entity data linking, self-identified bugs in Devin's own review

Impact / Damage

Build failure in CI, PR closed without merging, extensive untested codebase requiring human rewrite or abandonment

Share this incident

Help others know about this AI agent failure

Source: Github Pr View source Reported March 22, 2026

Frequently Asked Questions

What happened in incident STUPID-2026-0021?

Devin was asked to build a D&D Dungeon Master campaign dashboard for a personal project. Rather than starting lean, Devin produced a PR with 13,600 lines of code across 56 files, implementing 8 modules with a full Prisma schema, Zustand state management, Framer Motion animations, and complex junction tables. The Vercel deployment immediately failed. Devin's own review tool flagged 4+ issues in the code it had just written. ESLint rules were deleted from config rather than addressed. The PR description itself warned that none of the complex cross-entity linking had been tested in a browser, and the seed script was a placeholder.

Which AI agent caused this failure?

Devin was responsible for this scope explosion incident, documented as STUPID-2026-0021 in the StupidLLM AI agent incident database.

How severe was this AI agent failure?

It is rated 5/10 (medium) on StupidLLM's CVSS-style severity scale for AI agent failures, based on damage type, reversibility, and scope.

What was the root cause?

The root cause was classified as instruction misunderstanding. A functional, testable campaign dashboard with core modules

What was the impact or damage?

Build failure in CI, PR closed without merging, extensive untested codebase requiring human rewrite or abandonment