本文へ移動
cccskills
無料GitHub で公開

expense-review-policy

Review invoices and contracts against accounts-payable policy before human approval.

インストール方法を見る

含まれるファイル(1)

  • SKILL.md2.6 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

Expense review policy

Policy ID: AP-104.

Apply this policy to invoices, expense receipts, and service contracts.

Invoices and expenses

  • Read the entire invoice, including the vendor and every line item.
  • Extract every line item's quantity and unit price before calculating. Check the extracted list against every source line; an empty list is an error, not a zero subtotal.
  • Use Python's decimal.Decimal to multiply quantities by unit prices and sum the line totals. Compare this subtotal with the printed subtotal, then add shipping to calculate the total.
  • Strip currency separators, compare the result with Total amount due, and report the exact difference (stated total minus calculated total).
  • Require a purchase order or another documented approval.
  • Flag missing receipts, unsupported charges, and changed payment instructions.
  • Escalate incorrect totals and unverified changes to bank details.

Contracts

  • Read every clause and report all matching risks, even when one already requires escalation.
  • Flag automatic renewals and restrictive cancellation windows.
  • Flag unilateral price increases and missing liability limits.
  • Escalate customer-data sharing or subcontractor access without approval.
  • Identify missing confidentiality, security, or termination terms.

Review decision

  • needs_info: Required documentation or supporting information is missing.
  • escalated: Fraud indicators, financial discrepancies, or risky terms need review.
  • ready_for_approval: No policy violations remain, but a human must still approve.

Write /workspace/output/<document-stem>.json with the following fields:

{
  "document": "invoice.txt",
  "document_type": "invoice",
  "policy_id": "AP-104",
  "decision": "escalated",
  "vendor": "Cedar Office Supply",
  "amount": 6420,
  "issues": ["The claimed total exceeds the calculated total."],
  "recommendation": "Escalate for human review.",
  "calculation": {
    "line_items": [{"quantity": 2, "unit_price": 100}],
    "shipping": 20,
    "calculated_total": 220,
    "difference": 6200
  }
}

Every issues entry must be a plain-English string, not a nested object. Use the actual document values, not the illustrative numbers above. Include every invoice line item in calculation; amount is the stated total. For contracts, set document_type to contract and calculation to null.

Only the coordinator writes summary.json. A specialist writes its assigned report.

Never approve payments, sign contracts, or take external actions.

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

Bootstrap a new realtime eval folder inside this cookbook repo by choosing the right harness from examples/evals/realtime_evals, scaffolding prompt/tools/data files, generating a useful README, and validating it with smoke, full eval, and test runs. Use when a user wants to start a new crawl, walk, or run realtime eval in this repository.

日本語の概要は準備中です。原文の説明を表示しています。

openai/openai-cookbook7.7万2026年10月10日 更新

Review and edit OpenAI Cookbook notebooks, Markdown, and MDX for technical accuracy, clarity, grammar, consistency, runnable examples, and repository publication requirements. Use for editorial reviews, pre-merge documentation checks, or notebook Markdown-cell sweeps in openai-cookbook.

日本語の概要は準備中です。原文の説明を表示しています。

openai/openai-cookbook7.7万2026年10月10日 更新

openai のスキルをすべて見る

このスキルの問題を報告する