Changelog

Bug-fix release addressing multi-file init, check generation, and Windows terminal rendering.

• fixFixed multi-file init header showing only first file path instead of all initialised files.
• fixMerged baselines now generate 24 quality checks (was 18) — std_dev and unique checks are correctly preserved through the merge step.
• fixFixed middle-dot separator rendering on Windows terminals (replaced Unicode middle dot with ASCII pipe).
• fixFixed datetime sample value truncation in the profile table — full ISO timestamp now displayed.
• fixRemoved unused import; reporter version string now reads from the installed package version rather than a hardcoded constant.

Major feature release. Multi-file baselines, named profiles, 3 new check types, HTML reports, and scan history.

• newMulti-file baseline: pass multiple files to init — pipedog init jan.csv feb.csv --profile sales
• newNamed profiles: --profile flag on both init and scan lets you track multiple file types independently with one tool.
• new3 new check types: row_count (80% threshold), allowed_values (string/boolean columns with ≤ 50 distinct values), std_dev_change (distribution shift > 50%).
• newHTML reports auto-saved after every scan to .pipedog/<profile>/reports/. Self-contained — no server required.
• newScan history: every scan appends a record to history.json, creating an audit trail.
• newchecks list, checks edit, checks add subcommands for managing quality rules without touching JSON directly.
• newreport command to list saved HTML reports and open them in the default browser.

Initial public release.

• newpipedog init, scan, and profile commands.
• newCSV, Parquet, and JSON support out of the box.
• new5 baseline check types: not_null, null_rate, min_value, max_value, unique.
• newSchema drift detection: added/removed columns and type changes flagged automatically.
• newRich colour-coded terminal output powered by Rich.