Opinionated conventions for clean, reproducible and fast pandas data analysis in Python.
Profiles CSV and spreadsheet exports before you trust them: infers column types and measures missing values, duplicates, mixed types, outliers, date format chaos, and key uniqueness with a tested Python script, then writes a prioritized data quality report with safe fixes.
Checks CSV time series (metrics, sensor readings, sales by hour) for data problems before you chart or alert on them: missing intervals, duplicate or out-of-order timestamps, empty values, flatlines from stuck exporters, robust median/MAD spikes, and counter resets. Use when a dashboard looks wrong, before training a forecast, or when asking "is this data complete?". Includes a stdlib Python checker with --self-test.
Compares two CSV exports of the same records (old vs new, billing vs ledger, source vs warehouse) by key after normalizing headers, delimiters, currency formats, accounting negatives, dates and case. Reports added, removed and duplicate keys, key format drift, changed values per column with tolerances, possible re-keys and a control-total bridge, then writes a reconciliation report. Includes a stdlib Python checker with --self-test.
Finds what you cannot see in pasted code, configs, prompts and data files - Trojan Source bidi controls, hidden Unicode tag text (prompt injection), zero-width and invisible characters, odd spaces, control characters, Cyrillic/Greek homoglyphs, mojibake, invalid UTF-8, BOMs, mixed line endings and non-NFC text. Decodes hidden text, ranks findings and writes a cleaned copy. Includes an offline stdlib Python checker with --self-test.
Watches your training runs: summarises loss curves and metrics, detects divergence, NaNs and stalled GPUs early, and writes a comparison when runs finish.
Tracks SLO error budgets across services and tells you when burn rate means it is time to freeze risky deploys.
