Project ESPAÑOL
Find Spanish poems matched to your grammar learning level
Visit dmerullo.github.io ↗What is Project ESPAÑOL?
Project ESPAÑOL is an interactive tool for Spanish learners to discover classical poetry suited to their current skill level. It indexes roughly 10,000 public-domain poems spanning five centuries and 200 authors, computationally analyzed by verb tense frequency and grouped into beginner through expert difficulty tiers. Users filter by grammatical features, time period, and author nationality, then explore individual poems with their full grammatical profiles.
The creator built the initial version by hand in late 2024, using large language models to construct a Plotly Dash application. A recent update rewrote the front-end in JavaScript and Plotly.js with Claude's help in under an hour. The tool answers specific queries like identifying which 19th-century poets favored particular verb forms.
Momentum
Daily activity over the last 30 days: upvotes, unique visits, new GitHub stars and Hacker News points.
How do we know it's vibe coded?
vibetrend discovered Project ESPAÑOL automatically. Evidence: HN post: "using LLMs primarily to make an interactive Plotly Dash application". View source.
Similar vibe-coded apps

Interactive tool for exploring scales, chords, and harmony across instruments

Learn AI-powered coding by cloning Google, YouTube, Twitter, Reddit, and Amazon

A small programming language built on pure, untyped lambda calculus

Spaced-repetition trainer for systems performance and cost napkin-math numbers




Math language that compiles to C, Python, Verilog, and Lean theorems