Skip to main content
Glama
earl-hash

C4C: Mortgage Industry Regulations

by earl-hash

C4C: Mortgage Industry Regulations

A read-only MCP server that gives your AI assistant the verbatim text of U.S. mortgage regulations, with citations and effective dates, so it quotes the rule instead of guessing.

From Claude for Compliance.

This repository holds the public documentation for the hosted server: the tool list, example prompts and the registry server.json. The server itself is hosted; there is nothing to build or run locally.

What it covers

HUD Handbook 4000.1 (FHA), the Fannie Mae Selling and Servicing Guides, the Freddie Mac Seller/Servicer Guide, the Ginnie Mae MBS Guide, the VA Lenders Handbook (M26-7) and Servicer Handbook (M26-4), USDA HB-1-3555, the SBA SOP 50 10, plus the CFR and U.S. Code sections they rest on (Regulations B, X and Z, and others).

Every answer carries:

  • the verbatim regulator text, never a paraphrase presented as a quote

  • the citation and the effective date

  • the dated source document it came from

  • a currency flag when it applies: two versions in force, a pending revision, superseded, expired, or changed by a later letter or announcement

Related MCP server: Cal MCP Server

What you can do with it

  • Ask in plain words or loan-file shorthand ("DTI", "RSU income", "IRRRL seasoning"). The assistant finds the governing section and quotes it.

  • Look up a citation you already have: B3-4.3-04, 4000.1 II.A.4.c, 12 CFR 1026.43, 38 U.S.C. 3709, M26-7 Chapter 4 Topic 9.

  • Check whether a passage, including one from your own draft, really appears in the regulator's text (verify_quote).

  • Run guided playbooks in the same chat: FHA, VA and USDA loan-file QC, and a review against Fannie Mae Lender Letter LL-2026-04 (AI). Your documents stay in your own chat; the server never receives them.

See TOOLS.md for the ten tools and five prompts, and examples/prompts.md for starter prompts.

Install

Claude (web, desktop, mobile): Settings → Connectors → Add custom connector → paste the server URL. Available on any Claude plan (the free plan allows one custom connector).

ChatGPT: Settings → Apps / Connectors → create a connector with the server URL. ChatGPT allows connectors on Plus, Pro, Business, Enterprise and Edu (not its free plan). Claude for Compliance itself is free.

Claude Code:

claude mcp add --transport http c4c-mortgage-regs https://mcp.claudeforcompliance.com/mcp

Cursor / VS Code / other MCP clients:

{ "name": "c4c-mortgage-regs", "type": "http", "url": "https://mcp.claudeforcompliance.com/mcp" }

Privacy, limits and support

  • Read-only. No tool writes anything.

  • No accounts, no IP addresses stored. We count usage by AI platform only (for example "claude" or "chatgpt"), never by person, and keep the text of searches that found nothing so we can fill corpus gaps. Don't put borrower information in a search. Details: https://claudeforcompliance.com/privacy/

  • No sign-in, by design. The data is public regulatory text.

  • Limits: a per-IP rate limit and a daily request budget protect the free service.

  • Not legal advice. Results are for research. Verify against the primary source. Terms: https://claudeforcompliance.com/terms/

  • Support: earl@cfsynergies.com

License

The contents of this repository (documentation and examples) are MIT licensed; see LICENSE. The hosted service, its corpus and its source code are not part of this repository and are not covered by that license. Regulatory text belongs to its issuing agencies.

Related MCP Connectors

Related MCP Servers

  • A
    license
    A
    quality
    B
    maintenance
    Enables AI assistants to answer mortgage-related queries by providing tools for lender search, loan limit lookup, down-payment assistance programs, and more, with data sourced from real wholesale lenders and broker-curated intel.
    8
    8 npm
    MIT
  • A
    license
    Not graded
    quality
    D
    maintenance
    Enables AI assistants to parse and analyze mortgage documents (Loan Estimates & Closing Disclosures), converting them into structured MISMO-compliant JSON and checking for TRID compliance violations.
    2
    MIT