ClarityKit

• 

  main page

Inspiration

Opportunity pages often pack important details into long rules: age and student eligibility, team limits, exact submission items, AI-use disclosure, deadlines, and disqualification clauses. Those details are easy to overlook when someone is excited to start building. We wanted to make the next action clearer without asking people to trust an unexplained AI-generated answer.

What it does

ClarityKit accepts pasted text from an official notice and creates an evidence-backed readout. It flags likely deadlines, eligibility terms, required deliverables, restrictions, project-development rules, judging criteria, and prize language. Each finding includes the exact source wording that triggered it. A date panel surfaces common date formats; a checklist helps users verify important items; category filters narrow the readout; and CSV export makes the findings portable.

How we built it

ClarityKit uses Python and Flask for the app and JSON endpoint, regular expressions and keyword rules for explainable text extraction, and HTML, CSS, and vanilla JavaScript for the responsive interface and interactions. It does not require a paid API key or send the pasted notice to an external AI service.

Challenges we ran into

Rules are written inconsistently. A keyword can appear in an informational sentence rather than a strict requirement, and some deadlines use unusual date formats or time zones. ClarityKit therefore treats matches as prompts to verify, includes the triggering wording, and avoids making definitive eligibility or legal claims.

Accomplishments that we're proud of

ClarityKit is intentionally evidence-first: it shows the source text behind a match rather than presenting a mysterious risk score. It also explicitly labels its limits and reminds users to verify every item against the original source. This keeps the prototype useful without pretending that simple keyword matching is perfect.

What we learned

A helpful tool should make it easier to check the source, not replace the source. Transparent evidence and explicit uncertainty can be more useful than a confident-looking score that is hard to validate.

What's next for ClarityKit

Improve date/time-zone parsing, add configurable rules for different notice types, evaluate precision and recall on a labeled set of notices, and add accessible support for PDF or webpage import only after implementing safe parsing and source attribution.

Built With

• css
• flask
• html
• javascript
• python
• regular-expressions
• unittes