SuperAPI AI Directory
Partner with us中文AI newsFavorites
Skillsreceipt-expense-workflow

receipt-expense-workflow

Office & productivity

View repository

Updated 10-15 16:46

About this skill

Receipt Expense Workflow

An automatic receipt sorting, renaming, and aggregation tool using OCR (Surya).

When to Use

  • Sort receipts, Organize receipts, receipt sorting
  • I want to automatically organize receipt images in a folder by project
  • I want to name a large number of receipt photos with date, amount, and purpose
  • I want to batch rename and aggregate business trip expense receipts

Core Workflow

  1. Receive images / PDF / ZIP / videos from the input folder
  2. Extract date, amount, country, card, and purpose via OCR
  3. Apply quality correction and normalize to standard file names
  4. Sort by project
  5. Output a Markdown summary

Architecture

入力フォルダ（incoming/unassigned 等）
  │
  ├─ JPEG/PNG/PDF/ZIP → OCR → 抽出 → リネーム → PJフォルダへ移動
  ├─ MP4/MOV → ffmpeg → JPG → OCR → ...
  │
  ├─ カード不明 or テキスト不可 → 未分類フォルダ → 再仕分け可能
  │
  └─ 個人経費（D365対象外） → PJフォルダ/private/

private/ Folder: Store personal expense receipts that are not submitted to D365. Move here when the user determines after OCR that it is a "personal expense".

Naming Convention

YYYY-MM-DD-国コード-金額[-カード]-分類.ext

例:
  2026-02-09-jpn-14320-amex-shinkansen.jpeg
  2026-02-11-us-272-amex-hotel.jpeg
  2026-02-12-us-190.38-meal.jpeg        ← カードなし
  2026-02-09-us-6.80-amex-meal-starbucks.jpeg

For the country code list, classification list, D365 category mapping, and attachment procedure, refer to references/d365-expense-guide.md .

CLI Usage

# Dry-run（確認のみ、ファイル移動なし）
python receipt_sorter.py --project "202602_Domestic_Trip" --dry-run

# 本実行
python receipt_sorter.py --project "202602_Domestic_Trip" --input "incoming\unassigned"

# ログ指定
python receipt_sorter.py --project "202602_Domestic_Trip" --input "incoming\unassigned" --log "result.csv"

Options

Options Description Default
--project Output PJ folder name (required) -
--input Input folder incoming/unassigned/
--dry-run Display results only without moving files false
--log Result CSV path <PJフォルダ>/csv/<PJ名>_dryrun.csv or _result.csv

Markdown Summary Output

After the actual run, <PJ名>_summary.md is automatically generated in the PJ folder.

Included content:

  • Project total (mixed currencies, reference value)
  • Total table by country code
  • Total table by date x country code
  • All line items table (date, country, amount, card, classification, Japanese summary, file name)

Intake Pattern

  • Unclassified raw files should be consolidated not in the workspace root but in a dedicated intake folder
  • Example: incoming/unassigned/
  • Files whose project is confirmed may be placed directly into the project folder without going through the intake folder
  • For travel statements whose project is unknown, match read-only approval records, schedules, and reservation records by date, section, and purpose before moving. If there is no match, do not guess --project and leave it unclassified for confirmation
  • For system-issued transportation receipts, use the service date / boarding date, not the display date or purchase date. Only when both the departure point and arrival point can be confirmed, append origin-to-destination to the end of the classification to distinguish round trips
  • Limit the execution target to the dedicated intake, and perform the actual run after checking the dry-run results
  • Handle dry-run copies and verification inputs in a temporary folder, and delete them after completion

Setup

→ references/setup-guide.md

Script Reference

D365 Expense Integration

Attaching OCR-processed files to D365 Expense is an important use case of this skill, but the category table and attachment procedure are separated from the main body.

  • Report creation / metadata editing
  • Editing an Expense line's Category / Description / Country / tax group
  • line-level receipt upload / replacement / verification
  • read-only JSON verification of report / line / receipt card
  • Avoiding stale display and active row misalignment in the D365 UI

For details, refer to references/d365-expense-guide.md .

Known Limitations

  • OCR accuracy depends on the Surya model (misreads occur with handwritten or heavily distorted receipts)
  • The maximum candidate amount is adopted, but individual line items cannot be extracted from receipts with multiple line items
  • Amount concatenation misread: OCR may combine digits across line breaks and return a huge integer (e.g., 14520 + 300 → 14520300). This is defended against with an upper-limit check (anything over 5 million is excluded), but it is not perfect
  • Only one intermediate frame is extracted from videos (optimal frame selection is not implemented)
  • Japanese kanji store names are dropped during ASCII conversion when slugified (katakana is romanized)
  • Windows PIL file lock: Image.open() holds the handle and shutil.move() が WinError 32 fails with it.with + img.copy() + retries are already handled

Done Criteria

  • Can be processed consistently from OCR extraction through renaming
  • Can be sorted by project using standard file names
  • Can explain how unclassified and private are handled
  • A Markdown summary is output
  • When uploading to D365, it can be attached according to the reference matching rules