ohayoOHAYO / SKILL 003
Check the CSV.
Before import.
Catch duplicate keys, missing required cells and invalid types with rules you choose. Keep leading-zero IDs intact and turn findings into a verifiable fix list.
Hold: 6 findings across 4 records.
Record 2: duplicate SKU, invalid integer, decimal and date.
Record 3: required SKU cell is empty.
Record 4: row has too few columns.
See the rules, findings and corrected example →EXPLICIT RULES. ORIGINAL FILE.
From an export
to an import check.
Choose the rules
Confirm exact column names, required cells and per-column unique keys. Select text, integer, plain decimal or YYYY-MM-DD date types. IDs remain text.
Inspect locally
Run against a regular UTF-8 CSV with an explicit delimiter. Reports omit cell values, retain column names and identify record numbers and physical line spans. No network requests or third-party dependencies.
Verify a corrected copy
Confirm replacements with the data owner. Preserve the original and rerun with fresh report paths. The checker does not rewrite data or execute imports.
Know what a pass means.
Who is this for?
Developers, operations teams and small sellers checking repeat CSV imports with known rules. It does not require an AI subscription or a desktop application.
Which CSV formats and limits are supported?
UTF-8 with optional BOM; comma, semicolon, tab or pipe; quoted commas and newlines. Up to 10 MiB, 20,000 data records, 100 columns and 64 KiB rules. All findings are counted; first 500 details are retained. Exceeding input limits rejects the result.
Will it fix my data or verify inventory?
No automatic repairs, business accounting, inventory truth or importer compatibility guarantees. A pass means only the supplied supported rules passed. XLSX, delimiter guessing, non-UTF-8 input, symlinks and composite keys are unsupported.
What do the exit codes mean?
0: rules passed. 1: findings to review. 2: unsupported input, rules, limits or output failure; no accepted result. Reports require new paths. Empty optional cells pass type checks; add a required rule when they must be filled.
What about privacy and licensing?
Input stays local. Reports omit cell values and paths but still reveal column names and duplicate locations. One purchaser can use and modify the package for personal and client work and share reports; package redistribution is excluded.
How is this different from a free deduplication tool?
This package checks explicit rules and reports duplicates without deleting records. Python’s standard csv module and spreadsheets remain useful free alternatives; this download adds a repeatable workflow, evidence and tested examples.