Catch lending-file mistakes before money moves: have an AI agent read the loan pack and cross-check amounts, collateral and income against the credit opinion so a person only reviews the mismatches: AI agent prompt

At the same Nov 5, 2026 NH NongHyup Bank / AgileSoDA demo covered by Seoul Economic Daily, the audit department’s “zero lending accidents” idea drew a large crowd. The agent reads hundreds of loan documents and cross-checks loan amounts, collateral and income data against credit review opinions to surface errors and omissions before funding. Staff still own the decision; the agent’s job is to find inconsistencies humans miss in long packs. Confirm any rollout against your bank’s own model-risk and credit policy before using it on live files.

Curated by AgentPlaybook Real Estate · How we make and check prompts

Legal & ComplianceLoan-file cross-check before fundingintermediate

The prompt

1. Pick a closed loan file with the credit opinion and the supporting pack (application, income, collateral, appraisal).
2. List the fields that must match across documents (loan amount, collateral value, income figures, borrower IDs) and the rule for what counts as a mismatch.
3. Ask an AI agent to extract those fields from each document with citations and produce a mismatch table only—no approve/deny recommendation.
4. Have a credit or audit person walk every flagged row against the source pages and record false positives.
5. Only after several clean dry runs on historical files, use the mismatch table as a pre-funding checklist; a person still signs the decision.

Open in ChatGPT Open in Claude Perplexity Download as agent skill (SKILL.md)

How to use it

Copy the prompt above.
Replace every [bracket] with your own details (or fill in the blanks on AgentPlaybook Real Estate, which remembers your answers for the next prompt).
Paste it into Document AI / bank AI agent (or Claude/ChatGPT with uploads), loan file PDF pack, credit-opinion template and review the answer before you use it.

Works with: Document AI / bank AI agent (or Claude/ChatGPT with uploads), loan file PDF pack, credit-opinion template, mismatch checklist for auditors

Real-world example: en.sedaily.com/finance/2026/11/05/ai-cuts-five-day-project-finance-rep

Featured in: AI prompts for writing SOPs, reports, policies and checklists

Before you use it: what to check

Already built into this prompt

Ask an AI agent to extract those fields from each document with citations and produce a mismatch table only—no approve/deny recommendation.
Have a credit or audit person walk every flagged row against the source pages and record false positives.
Only after several clean dry runs on historical files, use the mismatch table as a pre-funding checklist; a person still signs the decision.

Check yourself

This prompt works on text other people wrote (emails, reviews, messages or web pages). That text can hide instructions aimed at AI, such as "ignore your rules and…". Paste it as material to work on, and never let the AI act on requests inside it, like paying, sharing logins or changing records. How prompt injection works →
Remove names, contact details, account numbers and other personal data the task doesn't need before you paste, and check your AI tool's data and training settings.
Check names, dates, numbers and any claim against your own source material before you use the answer. AI can misread or invent details.

Using an AI agent? Download this prompt as a SKILL.md agent skill with these checks built in, or get the .zip for Claude.ai. How to install skills.

</> Embed this prompt on your site

Paste this HTML into your blog or website. It's a plain card with no tracking scripts that links back to this prompt.

<blockquote style="border:1px solid #e4e4ee;border-left:4px solid #0f6b50;border-radius:10px;padding:14px 16px;margin:16px 0;font-family:system-ui,sans-serif;max-width:640px"><p style="margin:0 0 6px;font-weight:700">Catch lending-file mistakes before money moves: have an AI agent read the loan pack and cross-check amounts, collateral and income against the credit opinion so a person only reviews the mismatches</p><p style="margin:0 0 10px;color:#5c6078;font-size:14px">At the same Nov 5, 2026 NH NongHyup Bank / AgileSoDA demo covered by Seoul Economic Daily, the audit department’s “zero lending accidents” idea drew a large crowd. The agent reads hundreds of loan documents and cross-checks loan amounts, collateral and income data against credit review opinions to surface errors and omissions before funding. Staff still own the decision; the agent’s job is to find inconsistencies humans miss in long packs. Confirm any rollout against your bank’s own model-risk and credit policy before using it on live files.</p><a href="https://agentic-library.netlify.app/prompt/al-628-catch-lending-file-mistakes-before-money-moves-have-an-ai-ag/?utm_source=embed&amp;utm_medium=referral&amp;utm_campaign=prompt_embed" style="color:#0f6b50;font-weight:600">Get the free AI prompt on AgentPlaybook Real Estate →</a></blockquote>

Want to make this prompt even better? Grade it with the free prompt grader or build your own with the AI prompt generator.

Get the free Solo Founder Stack pack

32 copy-paste agent prompts (PDF) plus 5 new prompts every week.

Get it free

More Legal & Compliance prompts

Compliance evidence packCompliance · intermediate Claim intake assistantClaims · intermediate Keep a running log of a contract negotiationNegotiation tracking · intermediate Review a draft agreement against your precedents and playbookAgreement review · advanced Prep an anti-money-laundering alert review with every source linkedAML investigations · advanced Process product complaints and flag the ones regulators need to hear aboutQuality complaints · intermediate

See all Legal & Compliance prompts →

Related prompts from other teams

Competitor watch reportMarketing · beginner · worked example Market size & trend reportResearch · beginner Weekly status reportOperations · beginner · worked example Write an SOP from a messy processOperations · beginner · worked example