void-ghost000/salt-map-lite

SALT-MAP Lite is a Python tool that checks CSV spreadsheet files for hidden inconsistencies before you use that data for lookups, joins, reports, or feeding it to an AI tool. Many spreadsheet problems are not loud errors that jump out at you: things like the number 123 stored as 00123 in one row and as 123 in another, extra trailing spaces on an ID like EMP-001, numbers accidentally stored as text, duplicate keys that only differ by formatting, missing values, or Excel error codes like N/A. A person skimming the file may not notice these, but automated pipelines and AI systems that read the data often break or produce wrong results because of them. The tool scans a CSV file, groups the individual problems it finds into root cause trees so related issues are shown together instead of as a flat list, and generates two kinds of reports: a Markdown file meant for a person to read, with summaries and a collapsible tree of findings, and a JSON file meant for machines, AI tools, or data pipelines to consume directly. It runs entirely on your own computer, so your CSV file is never uploaded anywhere. The project includes ready made scripts for Windows PowerShell that let you try it immediately on sample data or point it at your own file, plus manual Python commands for other setups. The tool also includes an optional integrity gate with three modes, off, warn, and enforce, that checks whether its own generated reports are internally consistent using a hash chain. The README is explicit that this checks the report generation process itself, not the correctness of your original data or the true cause of how it became inconsistent in the first place. The project is currently an early alpha version, 0.2.0-alpha.1, and does not yet handle Excel XLSX files directly, only CSV. It is licensed under Apache-2.0 and is described as a smaller offshoot of a larger, separate project called SALT or HSRAG-SAIT.