Data cleaning and Visualization Sonnet

› In city-311-2/ there is requests.csv, a messy year of city 311 service requests, and README.md describes every column and its known problems. Using only Node built-ins for processing (no installs): write a cleaning script that parses the CSV properly, normalizes dates, categories, neighborhoods, statuses and durations, drops exact duplicates, and sets aside impossible durations; print what it removed or fixed and why. Then answer the README's questions and build a report as separate files: city-311-2/report/index.html, city-311-2/report/data.json with the cleaned aggregates, and city-311-2/report/report.mjs that fetches data.json and draws the charts (a chart library from an https CDN is fine). Lead the report with your three most important findings, state the cleaning decisions in a short methods note, and keep report/index.html working when opened from its folder with relative paths only.