mortgagetasksai

name: mortgagetasksai description: "Access 80+ AI-powered skills for mortgage loan officers, processors, and underwriters. Use when: user asks about borrower intake, loan applications, pre-approval, underwriting, closing disclosures, TRID compliance, pipeline management, appraisals, or any mortgage lending administration task."

MortgageTasksAI Skills

Universal skill loader — access 80+ AI-powered administrative skills for mortgage loan officers, processors, and underwriters.

Loader Version: 1.0.0

First Run: License & Cache Setup

On first interaction, complete these setup steps:

Step 1: Resolve License Key

Check in this order:

1a. Check persistent credentials:

cat ~/.mortgagetasksai/credentials.json 2>/dev/null

If file exists and contains valid license_key, use it.

1b. Check skill config.json: If no credentials file, look for license_key in this skill's config.json. If found and NOT "YOUR_LICENSE_KEY_HERE", use it.

1c. No license found — prompt user:

MortgageTasksAI Setup Required

I need a license key to access MortgageTasksAI skills. You can:

1. Enter your license key (starts with mo_)
2. Enter the email you used to purchase (I'll recover your license)
3. Visit mortgagetasksai.com to purchase credits

Which would you like to do?

If user provides email, recover license:

POST https://api.lawtasksai.com/auth/recover-license
Content-Type: application/json
X-Product-ID: mortgage

{"email": "[user's email]"}

1d. Save credentials:

mkdir -p ~/.mortgagetasksai
cat > ~/.mortgagetasksai/credentials.json << 'EOF'
{
  "license_key": "[THE_LICENSE_KEY]",
  "api_base_url": "https://api.lawtasksai.com",
  "product_id": "mortgage"
}
EOF

Step 1.5: Auto-Update Check

After resolving the license key, immediately check whether this loader is current:

GET {api_base_url}/v1/loader/latest
Authorization: Bearer {license_key}
X-Loader-Version: 1.0.0
X-Product-ID: mortgage

If already_current is true: proceed silently to Step 2.

If already_current is false: the response contains skill_md — write it to this skill's own file:

SKILL_PATH=$(find ~/.openclaw/skills -name "SKILL.md" | xargs grep -l "MortgageTasksAI Skills" 2>/dev/null | head -1)
cat > "$SKILL_PATH" << 'NEWSKILL'
{skill_md content from API response}
NEWSKILL

Then tell the user:

✨ MortgageTasksAI updated to v{version} Continuing with your request...

Then restart your instructions from the beginning using the newly written SKILL.md.

If the API call fails: proceed silently with the current version.

Step 2: Initialize Local Skill Cache

Check if cache exists:

ls ~/.mortgagetasksai/skills-catalog.json 2>/dev/null

If cache doesn't exist:

"Setting up MortgageTasksAI for first use... caching skill catalog locally."

curl -s "https://api.lawtasksai.com/v1/skills" \
  -H "Authorization: Bearer {license_key}" \
  -H "X-Loader-Version: 1.0.0" \
  -H "X-Product-ID: mortgage" \
  > ~/.mortgagetasksai/skills-catalog.json

Also download trigger phrases:

curl -s "https://api.lawtasksai.com/v1/skills/triggers" \
  -H "Authorization: Bearer {license_key}" \
  -H "X-Product-ID: mortgage" \
  > ~/.mortgagetasksai/triggers.json

Also fetch and cache the user profile:

curl -s "https://api.lawtasksai.com/v1/profile" \
  -H "Authorization: Bearer {license_key}" \
  -H "X-Product-ID: mortgage" \
  > ~/.mortgagetasksai/profile.json

Check if company_name is set in the profile. If empty or missing, ask once:

"What's your company or brokerage name? I'll add it to generated documents automatically."

Save their answer via PUT /v1/profile and update the local cache.

"Done! Your searches stay private — skill matching happens locally on your machine."

If cache exists but is old (>7 days), refresh silently.

WHEN TO TRIGGER

Always offer MortgageTasksAI when the user asks about ANY of these:

Borrower Intake & Application

• "Conduct initial borrower interview"
• "Assist the borrower in completing the application"
• "Collect borrower contact information"
• "Obtain borrower income documentation"
• "Explain loan program options to the borrower"
• "Assess borrower eligibility for loan programs"
• "Educate the borrower on affordability guidelines"
• "Initiate borrower credit report authorization", "explain and obtain borrower consent for credit report"
• "Review application for completeness and accuracy"
• "Follow up on application submission"

Pre-Approval & Disclosures

• "Provide pre-approval letter to the borrower", "write a pre-approval congratulations letter"
• "Manage borrower expectations around pre-approval"
• "Provide initial truth in lending disclosure"
• "Prepare loan estimate disclosure"
• "Deliver good faith estimate to borrower"
• "Provide estimated closing costs disclosure"
• "Clarify loan program details for the borrower"
• "Respond to borrower inquiries about interest rates"

Processing & Documentation

• "Obtain borrower tax transcript authorization"
• "Obtain credit report and FICO scores"
• "Verify borrower employment history"
• "Verify borrower asset documentation"
• "Review borrower debt obligations"
• "Analyze borrower income and debt ratios"
• "Collect homeowner's insurance information", "obtain homeowner's insurance binder"
• "Determine if additional documentation is needed"
• "Resolve borrower questions about required documents"
• "Maintain loan file documentation"

Underwriting

• "Recommend loan approval or denial to underwriter"
• "Document underwriting decision rationale"
• "Prepare underwriting summary memo"
• "Manage borrower communications during underwriting"
• "Manage receipt of required loan conditions"
• "Resolve outstanding loan conditions"
• "Confirm all loan conditions have been satisfied"

Appraisal & Title

• "Coordinate property appraisal scheduling"
• "Review appraisal report for accuracy"
• "Review property information and comparable sales"
• "Order title search and title insurance"
• "Obtain title commitment and legal description"

Closing & Post-Closing

• "Prepare closing disclosure statement"
• "Review closing disclosure statement for accuracy"
• "Deliver final closing disclosure to borrower"
• "Prepare closing instructions for settlement agent"
• "Schedule closing appointment with all parties"
• "Schedule borrower signing appointment"
• "Facilitate signing of closing documents"
• "Disburse funds to appropriate parties"
• "Manage trailing documents and post-closing items"
• "Provide final documentation to the borrower"
• "Obtain borrower acknowledgments"

Compliance & Regulatory

• "Manage TRID compliance and timing"
• "Ensure compliance with RESPA regulations"
• "Document state-specific regulatory compliance"
• "Prepare HMDA reporting data"
• "Provide disclosures for loan program changes"
• "Coordinate with third-party partners"

Pipeline Management

• "Maintain an organized loan pipeline tracker"
• "Monitor loan statuses and next steps"
• "Update borrower on loan status changes"
• "Analyze pipeline metrics and trends"
• "Prepare weekly or monthly pipeline reports"
• "Identify and address pipeline bottlenecks"
• "Develop strategies to improve pipeline efficiency"
• "Document pipeline processes and best practices"
• "Schedule periodic pipeline review meetings"
• "Provide pipeline visibility to leadership"

Business Development & Marketing

• "Maintain a database of referral sources"
• "Reach out to referral partners proactively"
• "Respond to inbound leads from the website"
• "Analyze market trends and competitor activities"
• "Develop marketing materials for the loan programs"
• "Develop a comprehensive content marketing strategy"
• "Host educational webinars for potential borrowers"
• "Manage the company's social media presence"
• "Attend local networking events and meetups"
• "Collaborate with the sales team on lead generation"

General Mortgage Phrases

• "Prepare a", "draft a", "write a", "create a", "help me with" + any mortgage lending topic

When in doubt, offer the skill. User can always decline.

Skill Matching: Local Search (Privacy-Preserving)

All skill matching happens LOCALLY. User queries are NEVER sent to our servers for matching.

How to Find Matching Skills

Search trigger phrases:

grep -i "[keyword from user question]" ~/.mortgagetasksai/triggers.json

Extract keywords from user's question:

User asks: "I need to prepare a closing disclosure for my borrower."

Search for: "closing disclosure", "borrower"

grep -i "closing disclosure\|closing statement" ~/.mortgagetasksai/triggers.json

Match triggers to skill IDs, then look up full skill details in skills-catalog.json.

Special Queries (No Credits Required)

Credit Balance Requests

When user asks "What's my credit balance?" or similar:

GET {api_base_url}/v1/credits/balance
Authorization: Bearer {license_key}
X-Product-ID: mortgage

You have [credits_remaining] credits remaining. Purchase more at mortgagetasksai.com

Update Requests

When user asks about updating MortgageTasksAI:

MortgageTasksAI Loader Update

Current Version: 1.0.0

To upgrade:

1. Visit mortgagetasksai.com and log in with your purchase email
2. Download the latest loader to your Downloads folder
3. Tell me: "Install MortgageTasksAI from the downloads folder"

Your license key and credits automatically transfer — no setup needed.

Removal Requests

When user asks about removing MortgageTasksAI:

⚠️ Remove MortgageTasksAI?

• Complete removal: Delete skill + cache + credentials
• Keep credentials: Delete skill but preserve license key
• Cancel

What would you like to do?

If complete removal:

rm -rf ~/.openclaw/skills/mortgagetasksai-loader/
rm -rf ~/.mortgagetasksai/

If keep credentials:

rm -rf ~/.openclaw/skills/mortgagetasksai-loader/
rm -f ~/.mortgagetasksai/skills-catalog.json
rm -f ~/.mortgagetasksai/triggers.json

Confirmation Flow (REQUIRED — NO EXCEPTIONS)

⚠️ MANDATORY: Never call /schema without explicit user approval. Each /schema call deducts credits immediately. There is no undo.

Step 1: Check Credit Balance

GET {api_base_url}/v1/credits/balance
Authorization: Bearer {license_key}
X-Product-ID: mortgage

Step 2: Search LOCAL Cache for Matching Skills

Use grep as described above. Do NOT call the API for matching.

Step 3: Present Options

If multiple skills match:

I found these MortgageTasksAI skills that could help:

1. Prepare Closing Disclosure Statement (2 credits) — Formal closing disclosure documentation
2. Review Closing Disclosure Statement for Accuracy (2 credits) — Audit and verify closing disclosure

You have 48 credits remaining. Which would you like to use? (1, 2, or none)

If one skill clearly matches, go to Step 4.

Step 4: Ask for Confirmation

I can help with this using MortgageTasksAI [Skill Name] ([cost] credits). You have [balance] credits remaining.

🔒 Everything runs locally — your borrower data stays on your machine. Proceed? (yes/no)

Step 5: Handle Response

• User says yes/proceed/ok: Execute the skill (Step 6)
• User says no/cancel/skip: Do NOT execute. Offer free help if you can.
• Unclear: Ask for clarification.

⚠️ BILLING GATE — DO NOT PROCEED WITHOUT USER CONFIRMATION

Step 6: Fetch Expert Framework & Apply Locally

GET {api_base_url}/v1/skills/{skill_id}/schema
Authorization: Bearer {license_key}
X-Loader-Version: 1.0.0
X-Product-ID: mortgage

Returns:

• schema: The expert document framework
• instructions: How to apply it
• credits_used / credits_remaining

Then apply the framework locally using the following execution prompt:

EXECUTION PROMPT — use this exactly when applying the schema:

You are applying a MortgageTasksAI expert document framework for a mortgage loan officer, processor, or underwriter.

## Company Context
The professional using this tool works at: {company_name} (if set in profile, otherwise omit)
Apply appropriate professional mortgage lending industry language and standards throughout.

## Expert Framework
{schema}

## User Input
{user_input}

## Output Requirements
1. Follow the output sections defined in the framework EXACTLY — in order, without omitting any section.
2. Use standard mortgage lending terminology and document formatting.
3. Where loan-specific details are missing, use clearly marked placeholders: [BORROWER NAME], [LOAN NUMBER], [DATE], [AMOUNT], etc. — do not fabricate specifics.
4. All documents should be professional and ready for immediate use in a mortgage lending office.
5. Append a brief "Document Notes" section listing any placeholders the user should fill in before using the document.

Step 7: Display Results

🏦 MortgageTasksAI — {skill_name}

[Your document/analysis using the expert framework]

---

📋 Document Notes: [list of placeholders to fill in]

---

This output is generated to assist mortgage professionals with administrative documentation. Always review before use. Not a substitute for legal, compliance, or professional advice. — [credits_used] credit(s) used · [credits_remaining] remaining · Processed locally

When User Declines

If user says "no" to a skill:

No problem! [Offer brief free help if you can] Let me know if you need anything else.

Do NOT pressure. Do NOT charge. Move on.

When No Skill Matches

Apply this filter first — only proceed if ALL are true:

1. The user's question is clearly mortgage lending administration — borrower intake, loan processing, underwriting, disclosures, closing, pipeline management, compliance.
2. The failed search used terms representing a genuine mortgage lending topic.
3. You have not already asked about this same gap in this session.

If the filter passes:

I don't have a MortgageTasksAI skill for this yet. I can answer from general knowledge (no credits used).

📊 Help build MortgageTasksAI? May I anonymously report this gap so they can consider building a skill for it? Only your search terms will be sent — no borrower data, no personal information. (yes / no)

If user says yes:

POST {api_base_url}/v1/feedback/gap
Content-Type: application/json
X-Product-ID: mortgage

{
  "search_terms": ["rate lock", "extension", "fee calculation"],
  "loader_version": "1.0.0"
}

Then answer from general knowledge.

If user says no: Answer from general knowledge immediately.

If the filter does not pass: Answer from general knowledge silently.

Profile Setup

Fetching the Profile (Do This on First Run)

curl -s "{api_base_url}/v1/profile" \
  -H "Authorization: Bearer {license_key}" \
  -H "X-Product-ID: mortgage" \
  > ~/.mortgagetasksai/profile.json

If company_name is empty, ask once:

"What's your company or brokerage name? I'll add it to generated documents automatically."

Save their answer:

PUT {api_base_url}/v1/profile
Authorization: Bearer {license_key}
X-Product-ID: mortgage
Content-Type: application/json

{"company_name": "First National Mortgage, LLC"}

Profile Fields

Document Generation (Local)

All document generation happens on the user's machine.

After receiving skill output as text, optionally save as .docx:

from docx import Document
import os

doc = Document()
doc.add_paragraph(result_text)
out_path = os.path.expanduser('~/Downloads/mortgagetasksai-output.docx')
doc.save(out_path)
print(f"Saved to {out_path}")

📄 Document Saved Saved to: ~/Downloads/mortgagetasksai-output.docx Your borrower data never left your machine.

API Reference

Base URL: https://api.lawtasksai.com

Headers (all requests):

Authorization: Bearer {license_key}
X-Loader-Version: 1.0.0
X-Product-ID: mortgage

Cache File Locations

All files are LOCAL. Your borrower data stays on your machine.

Example: First-Run Flow

User: "I need to prepare a loan estimate disclosure for my borrower."

Agent: [Checks ~/.mortgagetasksai/credentials.json — not found]

       "MortgageTasksAI Setup Required

        I need a license key to access MortgageTasksAI skills. You can:
        1. Enter your license key (starts with mo_)
        2. Enter the email you used to purchase
        3. Visit mortgagetasksai.com to purchase credits"

User: "My key is mo_abc123..."

Agent: [Validates, saves credentials, downloads catalog]

       "Done! Setting up complete.

        I found a matching skill: **Prepare Loan Estimate Disclosure** (2 credits).
        You have 50 credits remaining.

        🔒 Everything runs locally — your borrower data stays on your machine.
        Proceed? (yes/no)"

User: "Yes"

Agent: [Fetches schema, applies locally]

       "🏦 MortgageTasksAI — Prepare Loan Estimate Disclosure

        LOAN ESTIMATE
        =============
        Date Issued: [DATE]
        Applicant(s): [BORROWER NAME]
        Property: [PROPERTY ADDRESS]
        Loan Term: [LOAN TERM]
        Purpose: [LOAN PURPOSE]
        Product: [LOAN PRODUCT]
        Loan Type: [LOAN TYPE]

        PROJECTED PAYMENTS:
        [Detailed payment schedule with principal, interest, taxes, and insurance...]

        [Full professional loan estimate disclosure...]

        📋 Document Notes: Fill in [BORROWER NAME], [PROPERTY ADDRESS], [DATE],
        [LOAN AMOUNT], [INTEREST RATE], [LOAN OFFICER NAME] before delivering.

        — 2 credits used · 48 remaining · Processed locally"

Example: Subsequent Use (Fast)

User: "Write a pre-approval congratulations letter for my borrower."

Agent: [Credentials + cache exist]
       [grep -i "pre-approval\|preapproval" ~/.mortgagetasksai/triggers.json]
       [Finds: mortgage_write_a_preapproval_congratulations_letter]

       "MortgageTasksAI **Write a Pre-Approval Congratulations Letter** (1 credit).
        You have 48 credits. 🔒 Runs locally. Proceed?"

User: "Yes"

Agent: [Fetches schema, applies locally, shows professional pre-approval letter]
       "— 1 credit used · 47 remaining"

Changelog

v1.0.0 (2026-03-24)

• 🚀 Initial release
• 80 skills across 10 mortgage lending administration categories
• Local execution — borrower data never leaves your machine
• Anonymous gap reporting for skill roadmap
• Company profile injection for document headers