agentsclimarketplace
Sign in

Financial csv tool

Skill sammdu/financial-csv-tool/financial_csv_tool

General-purpose command-line transformer for financial and transaction CSV exports (bank, card, vendor billing). Use when an agent needs to filter, map, reshape, group, or conditionally expand statement rows into a target import schema.From its SKILL.md

Install
npx -y skills add sammdu/financial-csv-tool --skill financial_csv_tool

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

4.9 KB, ~1.2k tokens by cl100k_base, as published. Nobody here has run it

financial_csv_tool

A single PEP 723 self-contained uv script: scripts/financial_csv_tool.py. Run it from this skill directory with uv run scripts/financial_csv_tool.py INPUT [flags].

Flag reference

uv run scripts/financial_csv_tool.py --help is the authoritative flag list. Do not rely on this skill for per-flag details; it documents only what --help cannot convey.

Processing order

Transforms run in a fixed order regardless of flag order on the command line:

  1. Parse input: --skip-lines-start/--skip-lines-end, delimiter, cell stripping
  2. Exact-value row filtering with --where
  3. --group-by with --group and --group-collapse
  4. --rename
  5. --split
  6. --map
  7. --date
  8. Date-range filtering
  9. --coalesce, then --concat
  10. --expand, selected by --expand-by when present (also runs when --group-by is given without --group-collapse, laying out each group as a top entry with continuation rows)
  11. Column selection (--template, --keep, --drop) and --require checks

Consequences:

  • --group-by and --group reference original input column names; they run before --rename. The group key is tracked through renames for the expand step.
  • --split, --date, --coalesce, --concat, and --expand reference post-rename names and may consume columns created by earlier steps (a --concat source can be a --date output, an --expand amount can be a --split output).
  • --map "OUT=COL" "SOURCE=OUTPUT" ... maps every retained source value and fails on unmapped values.
  • --concat-sep "OUT=SEP" overrides the separator for one --concat output; repeat it for other outputs. Unspecified outputs use a space, and an empty SEP joins fields directly.
  • --expand-by COL selects rules written as VALUE=AMOUNT_COL[,COL...]=ACCOUNT:side; * selects every row.
  • --keep, --drop, and --require see the final post-transform columns.

Group and expand

  • --group-by COL reduces rows sharing a key. Repeat --group "OUT=COL[,COL...]" for derived values, choose the reducer with --group-type, and add --group-collapse when only the first row of each group should remain.
  • --expand turns one row into multiple lines. Without --expand-by, rules use AMOUNT_COL[,COL...]=ACCOUNT:side and apply to every row.
  • --expand-by COL selects rules by field value. Rules use VALUE=AMOUNT_COL[,COL...]=ACCOUNT:side; use * for a rule shared by every value.
  • Grouping and expansion may be combined: grouping establishes parent-entry boundaries and values, then expansion emits their child lines.

Skip-lines symmetry

Exports wrapped in metadata take --skip-lines-start N for lines before the real header and --skip-lines-end M for trailer lines after the data. Both count physical lines. Blank lines inside the data are dropped automatically and need no skipping.

Input and output

  • Input - (or omitting the argument) reads STDIN; output then goes to stdout unless -o is given.
  • A file input without -o writes <input stem>.transformed.csv next to the input.
  • Output reuses the input delimiter, so a tab-separated input stays tab-separated.

Keep transforms minimal

Preserve source columns and add as few new ones as possible. Only trim columns (--keep, --drop, --template) when the downstream importer requires a fixed header.

Stay context-agnostic

The tool is vendor-, business-, and importer-agnostic: all vendor-, business-, and importer-specific choices belong in command-line flags, never in its code. Usage of this skill is context-agnostic too; accounting treatment and target-system conventions belong to downstream skills.

Examples

Split a signed amount column into Deposit and Withdrawal by sign:

uv run scripts/financial_csv_tool.py statement.csv --split from=Amount to=Deposit,Withdrawal by=sign

Keep only rows with a specific transaction type:

uv run scripts/financial_csv_tool.py statement.csv --where "Type=Payment"

Map source values and select matching expansion rules:

uv run scripts/financial_csv_tool.py activity.csv --map "Entry Type=Type" "Charge=Journal Entry" "Payment=Bank Entry" --expand-by Type --expand "Charge=Amount=Expense:debit" --expand "Payment=Amount=Bank:credit"

Skip a 4-line preamble and a 2-line footer around the real header, normalizing the date column:

uv run scripts/financial_csv_tool.py export.csv --skip-lines-start 4 --skip-lines-end 2 --date "Date=Posted Date"

Verification

After transforming a financial statement, reconcile the output against the source: row count and per-amount-column sums must match to the cent. Report intentional exclusions, such as rows removed by date-range filtering or skipped lines, together with their totals.

What ships with it: 3 files

21.6 KB alongside SKILL.md, 1 of them executable

evals/

scripts/

Gives 0 of the 12 instructions most finance skills give in ~1.2k tokens

Counted across 469 of the 469 authors here whose files we hold, read 2026-08-07

  • Extract date vendor amount and descriptionin 15 of 469, across 3 files
  • Scan folder for invoice filesin 14 of 469, across 2 files
  • Rename files to standard formatin 14 of 469, across 2 files
  • Show organization plan before movingin 14 of 469, across 2 files
  • Generate summary CSVin 14 of 469, across 2 files
  • Organize files by categoryin 13 of 469, across 1 file
  • Preserve original filesin 13 of 469, across 1 file
  • Flag files missing critical infoin 13 of 469, across 1 file
  • Produce the requested output filein 9 of 469, across 4 files
  • Build best, base, and worst case scenariosin 9 of 469, across 5 files
  • Implement backoff if rate limit errors occurin 8 of 469, across 3 files
  • Determine the weighted average cost of capitalin 8 of 469, across 4 files

Said here and by no other author read

  • consult --help for authoritative flag details
  • preserve source columns and add few new ones
  • trim columns only when importer requires fixed header
  • keep vendor-specific choices in command-line flags only
  • leave accounting treatment to downstream skills
  • leave target-system conventions to downstream skills

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.