All toolsUIQA
UX checks before the merge.
An open-source CLI and GitHub Action for repeatable product-experience risks in frontend code.
Built by Insan NurjamanMIT licensed
npm install -D @insannurjaman/uiqa UX001
A list with nowhere to go
{items.map(item => (
<ProjectCard key={item.id} item={item} />
))}
- Review risk
- The collection has a populated view, but no visible path for zero items.
- Suggested next step
- Add an empty state that explains what happened and offers a useful next action.
Illustrative code example, not a scan of this portfolio. Read the rule reference
Code can pass.
The experience can still fail.
I built UIQA to bring repeatable UX review questions into the pull-request workflow. Empty collections, unavailable data, unclear errors, and weak recovery paths can survive a successful build.
The tool is a complement to engineering checks and human design review. It looks for inspectable implementation patterns, not whether a design is tasteful or whether every user journey works.
Why I built it Three layers of review.
- Product experience
- Empty, loading, error and no-result states; form feedback; destructive-action recovery; checkout cues; mobile layout risks.
- Accessible interaction
- Text alternatives for images and accessible names for interactive controls.
- Design-system consistency
- Hardcoded color and spacing patterns that can bypass shared tokens.
Explore the documented rules Small command.
Useful feedback.
Install as a development dependency, then scan the source directory. Reports include severity, category, file locations and suggestions.
The command below generates Markdown. JSON output is available for automation, and a severity threshold can make findings fail CI.
Full CLI documentation Install
npm install -D @insannurjaman/uiqaScan
npx uiqa scan --path ./src --format markdownUse in pull requests
- uses: insannurjaman/uiqa@v0.1.1
with:
path: ./src
format: markdown
fail-on: high
GitHub Action setup Deterministic.
Not a substitute for judgment.
UIQA is early v0.1 software. It performs static checks without AI or browser rendering. It does not compare screenshots or Figma designs, and it does not replace accessibility, usability, or product testing. Heuristics can miss issues or flag legitimate patterns.
SARIF, pull-request summary comments, broader framework detection and optional AI adapters are roadmap items, not capabilities claimed here.
Documentation and npm metadata checked October 8, 2026. Version 0.1.1 is the published version verified for the example workflow.
