AI coding agents hallucinate icon names and libraries: VALIDATE, 72/100 | Ideas Lab

A verification layer that catches AI-generated icon and library hallucinations before code execution.

The pain

Developers using AI coding agents waste 5-15 minutes per session debugging broken icon references and non-existent library imports. This breaks flow state and creates friction in the AI-assisted development workflow that's becoming standard.

The gap

AI models generate plausible-sounding but incorrect names because they optimize for fluency, not accuracy. No existing tool sits between the AI output and the developer to validate these specific hallucinations against real icon libraries and package registries.

Build angle

Start as a lightweight VS Code extension that real-time validates icon names (Material, Feather, Heroicons, etc.) and library imports against official registries. Surface warnings before code is committed. Expand to a broader AI output validator.

Strengths

Solves a concrete, repeatable problem that affects most developers using Claude/Cursor daily
Clear success metric: reduction in hallucination-related debugging time
Multiple revenue paths: extension sales, API licensing to IDE makers, integration into AI coding platforms
Low barrier to initial validation: build against 5-10 popular icon libraries first

Risks

Problem may be temporary if AI models improve hallucination rates in next 6-12 months
Market size unclear: how many developers actually use AI agents + hit this specific issue frequently enough to pay
Requires constant maintenance as libraries update and new icon sets emerge
Could be built as free feature by Cursor/Windsurf themselves, eliminating standalone business
Validation logic is straightforward to replicate; low defensibility

Fly Labs Method

Is the pain real, is there a gap, is it the right time, can one person build it.

72

Value Equation

Dream outcome and how likely it feels, against the time and effort it costs.

58

One-Person Business

Curiosity pull, identity fit, and a path from free value to paid for a solo creator.

68

Viral Frameworks

Hook strength, shareability, and how cheaply it can be tested.

75

Builder Lens

Evidence the problem exists, timing, defensibility, and a model that fits on a napkin.

64

Five lenses, one composite. How scoring works

The angle

Start as a lightweight VS Code extension that real-time validates icon names (Material, Feather, Heroicons, etc.) and library imports against official registries. Surface warnings before code is committed. Expand to a broader AI output validator.

No market research recorded for this idea yet.

Excessive CAPTCHA challenges degrading GitHub login UX for enterprise users

GitHub's anti-bot protection forces users, especially those with organizational emails, to solve 10 blurry puzzles multiple times, creating an extremely painful and frustrating user experience.

Dev

BUILD

86FL score

GTA modders tediously manually edit assembly files for modpacks

Creating .oiv modpacks for GTA is tedious and error-prone due to buggy project manager software, forcing manual editing of assembly files.

Dev

BUILD

85FL score

BreezePDF – Free, in-browser PDF editor

BreezePDF lets you edit, sign, merge, compress, redact, OCR, fill forms, extract tables, and use 30+ more PDF tools — all in the browser, no sign-up. Files never leave your computer.I built it because when people search Google for common PDF tasks, many of the tools they find upload documents to a server. I wanted an option that keeps files local instead.I posted an earlier version on HN last spring: https://news.ycombinator.com/item?id=43880962At the time it only supported a small set of features. Over the last 10 months I rebuilt large parts of it and expanded it to nearly 40 tools, including several ideas that came from comments in that earlier thread.There is also now a desktop app for macOS, Windows, and Linux, plus a CLI/SDK for developers.

Dev

BUILD

85FL score

All Dev ideas