DealMachine MCP server
Commit 13 Nov 2026Memungkinkan AI client untuk melakukan riset pemilik properti di AS, menemukan informasi kontak, menganalisis perbandingan penjualan, dan mengelola daftar prospek.
Juga dikenal sebagai DealMachine/dealmachine-cli, dealmachine-cli
- Stars
- 1
- Tools
- —
- Unduhan, 30 hari
- —
- Versi terbaru
- 0.3.1
- Commit terakhir
- 13 Nov 2026
- Lisensi
- MIT
Apa itu DealMachine MCP?
Memungkinkan AI client untuk melakukan riset pemilik properti di AS, menemukan informasi kontak, menganalisis perbandingan penjualan, dan mengelola daftar prospek.
Ringkasan dibuat dari README proyek
Tentang DealMachine MCP server
DealMachine menyediakan intelijen properti melalui command line interface dan MCP server yang dirancang untuk profesional real estate dan developer. Tool ini memungkinkan pengguna untuk meriset pemilik properti di AS, menemukan informasi kontak, dan menganalisis perbandingan penjualan. Tool ini dibangun untuk bekerja dengan AI client seperti ChatGPT, Claude, Cursor, dan GitHub Copilot menggunakan autentikasi OAuth 2.1 atau bearer-token via DealMachine API key.
Pengguna dapat mengelola daftar prospek, melakukan pengayaan data, dan memvalidasi alamat. Sistem ini menyertakan skill yang sadar kredit yang membantu agent menemukan filter dan field, menghitung potensi hasil, serta mengonfirmasi operasi berbayar skala besar sebelum dijalankan. Hal ini mencegah biaya tak terduga dengan memungkinkan AI untuk mengestimasi penggunaan kredit terlebih dahulu. Layanan ini tersedia melalui koneksi hosted MCP atau sebagai paket plugin agent portabel untuk client yang kompatibel.
Apa yang bisa Anda lakukan dengannya
- 1Identifikasi properti dengan ekuitas tinggiAI agent dapat menemukan properti milik pemilik yang tidak tinggal di lokasi tersebut (absentee-owned) dengan ekuitas tinggi di kota tertentu sambil mengestimasi biaya kredit terlebih dahulu.
- 2Melakukan riset pemilik propertiAnda dapat meminta AI untuk mencari pemilik dari properti tertentu dan mengambil detail kontak yang tersedia.
- 3Menganalisis perbandingan penjualanSeorang developer dapat menggunakan tool ini untuk menemukan data perbandingan penjualan untuk properti tertentu guna membantu proses valuasi.
- 4Membangun daftar prospek yang tertargetTool ini dapat meriset individu yang memenuhi kriteria spesifik untuk membuat daftar prospek baru.
Ditulis berdasarkan README proyek. Setiap use case diperiksa terhadap kutipan dari README tersebut.
Cara menginstal DealMachine MCP
Format diverifikasi 10 Nov 2026Tambahkan ke CursorTambahkan ke GooseInstal DealMachine MCP di Claude Code
$ claude mcp add --transport http dealmachine-cli https://mcp.dealmachine.com/Instal DealMachine MCP di Cursor
~/.cursor/mcp.json
{
"mcpServers": {
"dealmachine-cli": {
"url": "https://mcp.dealmachine.com/"
}
}
}Instal DealMachine MCP di VS Code
.vscode/mcp.json
{
"servers": {
"dealmachine-cli": {
"type": "http",
"url": "https://mcp.dealmachine.com/"
}
}
}Instal DealMachine MCP di VS Code CLI
$ code --add-mcp '{"name":"dealmachine-cli","type":"http","url":"https://mcp.dealmachine.com/"}'Instal DealMachine MCP di Codex
$ codex mcp add dealmachine-cli --url https://mcp.dealmachine.com/Instal DealMachine MCP di Codex config.toml
~/.codex/config.toml
[mcp_servers.dealmachine-cli] url = "https://mcp.dealmachine.com/"
Instal DealMachine MCP di Gemini CLI
$ gemini mcp add -s user --transport http dealmachine-cli https://mcp.dealmachine.com/Instal DealMachine MCP di Gemini settings.json
~/.gemini/settings.json
{
"mcpServers": {
"dealmachine-cli": {
"url": "https://mcp.dealmachine.com/",
"type": "http"
}
}
}Konfigurasi
DM_API_URL
| Nama | Deskripsi | Tipe |
|---|---|---|
| DM_API_URL | — |
Pengecekan otomatis
Berjalan 14 Nov 2026Laporan lengkap →Belum diverifikasi. Sebuah server memerlukan skor 80, tidak ada temuan serius, dan data yang cukup untuk kami periksa. Cara kerja checks · Yang perlu diperbaiki untuk menambah poin
- Source repository terhubung dan dapat dijangkaudealmachine/dealmachine-cli is public and reachablegithub.com/dealmachine/dealmachine-cli ↗Lolos
- Nama paket mencantumkan repositori ini sebagai sumbernyaTidak diperiksa
- Provenance build (atestasi Sigstore)Tidak diperiksa
- Lisensi open source (OSI)MITspdx.org/licenses/MIT.html ↗Lolos
- Ada commit atau release dalam satu tahun terakhirLast commit or release 2026-11-13github.com/dealmachine/dealmachine-cli ↗Lolos
- OpenSSF ScorecardTidak diperiksa
- Nama package tidak mirip dengan package populerTidak diperiksa
- Tidak ada advisory kritis pada paket atau dependensinyaTidak diperiksa
- Tidak ada instruksi tersembunyi untuk model0 tool descriptions and the README read; no override phrases, credential paths or hidden charactersLolos
- Tools yang bersifat destruktif telah diberi labelTidak diperiksa
- Endpoint remote merespons handshake MCPinitialize answered with protocol 2025-06-18 in 807 msLolos
README
Lihat di GitHubDealMachine CLI
Maintainers: agent instructions, development, current Command reference, and npm releases.
DealMachine CLI (dm) -- property intelligence from the command line.
A standalone Commander.js CLI that talks to the DealMachine REST API. Provides Commands for authentication, property and people research, enrichment, lists, prospects, tags, webhooks, mail and developer utilities. Compiles to ESM JavaScript via tsc.
This package has zero @dealmachine/* dependencies -- it is a self-contained binary that communicates exclusively through the public API.
AI agent integrations
This repository is also the public distribution package for the DealMachine MCP server and DealMachine skill.
- Hosted MCP server:
https://mcp.dealmachine.com - API documentation:
https://api.docs.dealmachine.com - Account and API keys:
https://dealmachine.com/settings/developer - Privacy policy:
https://dealmachine.com/privacy-policy - Terms of service:
https://dealmachine.com/terms-of-service - Support:
support@dealmachine.com
The MCP server supports OAuth 2.1 for ChatGPT, Claude, Cursor, Codex, and other compatible clients. It can also use a DealMachine API key in developer clients that support bearer-token configuration.
The plugin package includes:
- A hosted MCP connection for property, people, enrichment, comparable-sales, and account tools
- A credit-aware skill that discovers filters and fields, counts first, and confirms large paid operations
- A portable Agent Plugins package for compatible clients
- Manifests for OpenAI, Claude, Cursor, GitHub Copilot, and Gemini
- Official MCP Registry metadata in
server.json
The portable package follows Agent Plugins 1.0.0:
dealmachine-cli/
├── plugin.json
├── mcp.json
└── skills/
└── dealmachine/
├── SKILL.md
├── REFERENCE.md
└── SETUP.md
Compatible clients discover the DealMachine skill from skills/dealmachine/ and connect to the
hosted Streamable HTTP MCP server declared in mcp.json. Client-specific manifests remain in the
repository for compatibility, marketplace metadata, and richer presentation.
Example requests:
- "Find high-equity absentee-owned properties in Austin and estimate the credit cost first."
- "Look up the owner of this property and find available contact data."
- "Find comparable sales for this property."
- "Research people who match these criteria for a targeted prospecting list."
Direct skill installation:
npx skills add DealMachine/dealmachine-cli
Table of Contents
- AI agent integrations
- Installation
- Authentication
- Configuration
- Commands
- Agents --
agents,agents guide,agents playbook,agents install,agents permissions - Auth --
login,logout,whoami - Config --
config get,config set,config path - Account --
account - Usage --
usage - Properties --
search,count,get,ids,export - People --
search,count,get,ids,export - Enrich --
address,latlng,apn,email,phone,name - Comps -- comparable property analysis
- Lists --
search,create,get,update,delete,build,import,items,add,remove,export - Filters -- list available search filters
- Fields -- list available data fields
- Activity --
search,get - Addresses --
autocomplete,validate - Dev --
license add,license list,license remove
- Agents --
- Global Options
- Input Methods
- Project Structure
- Building
- Adding New Commands
- Dependencies
Installation
From npm (global)
npm install -g dealmachine
dm login
The canonical implementation package is @dealmachine/cli. The dealmachine package is the short install alias and provides the same dm command.
From source
cd dealmachine-cli
npm run build
node dist/index.js whoami
Link for local development
cd dealmachine-cli
npm link
dm --version
The binary entry is dist/index.js, declared in package.json under bin.dm. Requires Node.js >= 18.
Authentication
The CLI supports two authentication methods.
Device Auth Flow (RFC 8628)
The default dm login command uses the OAuth 2.0 Device Authorization Grant (RFC 8628). This is the recommended flow for interactive use:
dm login
- The CLI requests a device code from
POST /v1/auth/device/codewith client IDdealmachine-next-cliand your machine's hostname. - A verification URL and user code are displayed. The browser opens automatically (unless
--no-browser). - You authorize the device in the browser by entering the user code.
- The CLI polls
POST /v1/auth/device/tokenat the server-specified interval. - On success, the API key, key ID, and organization details are stored to
~/.dealmachine/config.json.
The polling handles all RFC 8628 responses: authorization_pending, slow_down (backs off by 5s), access_denied, and expired_token.
# Skip auto-opening the browser
dm login --no-browser
# Target a specific environment
dm login --env local
dm login --env staging
Direct API Key Login
For CI pipelines, scripts, or local development, pass an API key directly:
dm login --key dm_sk_live_abc123...
The key is verified against GET /v1/account before being stored. If verification fails, the CLI exits with a non-zero code.
If you do not have an API key yet, use dm signup <email> --login or dm login. Free accounts can count matches and estimate costs before paying. Public plan checkout only accepts self-serve Basic and Pro prices from the shared plan catalog and is capped at 60,000 monthly data credits.
Switching Environments
If you are already logged in, you can switch the target API environment without logging out:
dm login --env local # Switch to http://localhost:3001/v1
dm login --env staging # Switch to https://api-staging.v2.dealmachine.com/v1
dm login --env production # Switch to https://api.v2.dealmachine.com/v1
Logout
dm logout
Removes the config file at ~/.dealmachine/config.json.
Configuration
Credentials are stored at ~/.dealmachine/config.json with file permissions 0600 (owner read/write only). The config directory ~/.dealmachine/ is created with mode 0700.
Config File Schema
{
"apiKey": "dm_sk_live_...",
"keyId": "key_abc123",
"organizationId": 42,
"organizationName": "Acme Corp",
"organizationSlug": "acme-corp",
"apiEnvironment": "production"
}
Environment Variables
The CLI checks these environment variables for API URL resolution (in priority order):
| Variable | Purpose | Example |
|---|---|---|
DM_API_URL / DEALMACHINE_API_URL |
Direct URL override | http://localhost:3001/v1 |
DM_ENV / DEALMACHINE_ENVIRONMENT |
Environment name | local, staging, or production |
If none are set, the CLI falls back to the apiEnvironment field in the config file, then defaults to production.
API Environments
| Environment | URL |
|---|---|
local |
http://localhost:3001/v1 |
staging |
https://api-staging.v2.dealmachine.com/v1 |
production |
https://api.v2.dealmachine.com/v1 |
Commands
Agents Commands
dm agents
Print concise guidance for agents using the CLI. This is the recommended first command when an agent has access to dm but has not loaded the DealMachine Playbook yet.
dm agents
dm agents --json
The guide tells agents to use --json and --quiet, verify auth, fetch live filters and fields before searches, count before credit-consuming work, and confirm expected credit usage before fetching records or exporting.
dm agents guide
Print the same concise agent guidance explicitly.
dm agents guide
dm agents guide --json
dm agents playbook
Print the bundled DealMachine Playbook Markdown. Agents should load this before translating natural language property, people, contact, enrichment, list, export, comps, or credit-usage requests into CLI commands.
dm agents playbook
dm agents playbook --json
dm agents skill # alias
The public CLI source keeps its bundled Playbook at playbook/PLAYBOOK.md. The build writes the selected source to dist/agents/dealmachine-playbook.md, so the command works from a published CLI package as well as a local source checkout.
dm agents install claude-code
Install the Playbook as a native Claude Code skill. Personal scope is the default. Project scope installs under the current repository.
dm agents install claude-code
dm agents install claude-code --project
dm agents permissions
Print the narrow Claude Code allowlist for free discovery and count commands. Paid and mutating commands are not pre-approved.
dm agents permissions
dm agents permissions --json
Auth Commands
dm signup
Create a public API account and receive an API key. With --login, the credential is saved locally and omitted from output (including JSON); use this form from an agent:
dm signup developer@example.com --first-name Ada --last-name Lovelace --phone-number +15551234567
dm signup developer@example.com --login
dm plans
List live self-serve plans, included credits, features and free API access. Connected customers receive their workspace prices:
dm plans
dm plans --json
dm checkout
Preview the recurring terms using a current price ID from dm plans. After the customer approves the exact price, billing interval and quantity, add --yes to create the Stripe checkout link:
dm checkout --price-id price_xxx_monthly --quantity 1 --json
# Only after approval:
dm checkout --price-id price_xxx_monthly --quantity 1 --yes --json
Payment is completed on Stripe. Creating checkout does not activate a plan; check dm account --json and dm usage after payment before retrying enrichment. Counts do not award credits. Free operations still require authentication and obey permissions and rate limits.
dm login
Authenticate with your DealMachine account.
dm login # Device auth flow (opens browser)
dm login --no-browser # Device auth, manual code entry
dm login --key dm_sk_live_abc123 # Direct API key
dm login --env local # Target local API
| Option | Description |
|---|---|
--no-browser |
Do not automatically open the browser |
--key <api-key> |
Login directly with an API key (skips browser) |
--env <environment> |
API environment: local, staging, or production |
dm logout
Remove stored credentials.
dm logout
dm whoami
Show current authentication status.
dm whoami # Show stored credentials
dm whoami --verify # Verify credentials against the API
| Option | Description |
|---|---|
--verify |
Verify credentials with the API |
Config Commands
dm config get [key]
Get a configuration value, or display all values when no key is given.
dm config get # Show all config values
dm config get apiEnvironment # Show specific value
dm config get apiKey # Shows truncated key (first 20 chars)
Available keys: organizationName, organizationSlug, organizationId, apiEnvironment, keyId, apiKey.
dm config set <key> <value>
Set a configuration value. Only apiEnvironment is editable.
dm config set apiEnvironment local
dm config set apiEnvironment staging
dm config set apiEnvironment production
dm config path
Print the absolute path to the config file.
dm config path
# /Users/you/.dealmachine/config.json
Account Commands
dm account
Display account information including organization name, ID, creation date, and auth type.
dm account
Output:
Account
────────────────────────────────────────
Organization: Acme Corp
Org ID: 42
Created: Jan 15, 2025
Auth Type: api_key
Usage Commands
dm usage
Show credit usage for the current billing cycle.
dm usage # Human-readable table
dm usage --json # Machine-readable JSON
Output:
Credit Usage
──────────────────────────────────────────────────
Plan: Pro
Cycle: Mar 1, 2026 : Mar 31, 2026
Credits: 4,200 / 10,000 (42%)
Remaining: 5,800
Breakdown:
Properties: 3,100
People: 1,100
Properties Commands
dm properties search
Search properties with filters and locations.
# Inline JSON body
dm properties search --body '{
"locations": [{"type": "zip_code", "code": "78704"}],
"filters": [{"filter_id": "property_type", "operator": "is_any_of", "value": ["single_family"]}]
}'
# From a file
dm properties search -f search.json
# Pipe from stdin
cat search.json | dm properties search
# Machine-readable output
dm properties search -f search.json --json # Free estimate for scripts and agents
dm properties search -f search.json --json --yes # Run after approval
# Explicit free estimate
dm properties search -f search.json --estimate-cost
# Query Builder protocol filters
dm properties search --include-lists 123,456 --exclude-previously-exported --body '{"locations":[]}'
| Option | Description |
|---|---|
--body <json> |
Request body as JSON string |
-f, --file <path> |
Read request body from a JSON file |
--include-lists <ids> |
Comma-separated list IDs to include |
--exclude-lists <ids> |
Comma-separated list IDs to exclude |
--exclude-previously-exported |
Exclude records already exported by your organization |
--bigquery-data-environment <n> |
Query Builder data environment (1 production, 2 staging, 3 development) |
--estimate-cost |
Preview counts and credit cost without consuming credits |
--yes |
Confirm approved credit spend for non-interactive execution |
--json |
Output as JSON |
dm properties count
Count properties matching filters without consuming credits.
dm properties count --body '{"locations": [{"type": "state", "code": "TX"}]}'
dm properties count -f filters.json --json
dm properties get <id>
Get a single property by its DealMachine ID.
dm properties get prop_12345
dm properties get prop_12345 --contact-audience owners_and_family
dm properties get prop_12345 --contact-audience none
dm properties get prop_12345 --fields estimated_value,equity
dm properties get prop_12345 --json
| Option | Description |
|---|---|
--contact-audience <audience> |
owners, owners_and_family, renters, residents, all, none |
--fields <csv> |
Comma-separated property field IDs from dm fields |
--json |
Output as JSON |
Property lookup defaults to owners. If you only need property data, use --contact-audience none. This omits contacts and avoids people credits.
dm properties ids [ids...]
Get multiple properties by their IDs in a single batch request.
# Positional arguments
dm properties ids prop_111 prop_222 prop_333
# Via JSON body
dm properties ids --body '{"ids": ["prop_111", "prop_222"]}'
# From file
dm properties ids -f ids.json --contact-audience owners
dm properties ids -f ids.json --contact-audience none
| Option | Description |
|---|---|
--body <json> |
Request body as JSON string |
-f, --file <path> |
Read request body from a JSON file |
--contact-audience <audience> |
Include contacts: owners, owners_and_family, renters, residents, all, none |
--json |
Output as JSON |
dm properties export
Export properties as CSV (up to 1,000,000 records). Returns signed download URLs.
dm properties export -f search.json
dm properties export -f search.json --require-phone --scrub-dnc
dm properties export --body '{"locations": [...]}' --mobile-only --json
| Option | Description |
|---|---|
--body <json> |
Request body as JSON string |
-f, --file <path> |
Read request body from a JSON file |
--require-phone |
Only include records where the contact has a phone number |
--require-email |
Only include records where the contact has an email address |
--mobile-only |
Only include wireless phone numbers |
--landline-only |
Only include landline phone numbers |
--scrub-dnc |
Exclude contacts on the Do Not Call registry |
--json |
Output as JSON |
People Commands
dm people search
Search people with filters and locations.
dm people search --body '{
"locations": [{"type": "zip_code", "code": "78704"}],
"filters": [{"filter_id": "age", "operator": "between", "value": [30, 50]}]
}'
dm people search -f people-search.json --json
dm people search -f people-search.json --estimate-cost
dm people search -f people-search.json --json --yes
dm people search --include-lists 123 --exclude-lists 456 --exclude-previously-exported --body '{"locations":[]}'
Non-interactive People Search returns a free estimate unless --yes is supplied. A specific person
by name uses dm enrich name, not People Search.
dm people count
Count people matching filters without consuming credits.
dm people count -f filters.json
dm people get <id>
Get a single person by their DealMachine ID.
dm people get per_12345
dm people get per_12345 --include-properties --property-limit 20
dm people get per_12345 --fields estimated_household_income,estimated_value
dm people get per_12345 --json
| Option | Description |
|---|---|
--include-properties |
Include associated properties |
--property-limit <n> |
Maximum associated properties to return, from 1 through 100 |
--fields <csv> |
Comma-separated field IDs from dm fields |
--json |
Output as JSON |
dm people ids [ids...]
Get multiple people by their IDs in a single batch request.
dm people ids per_111 per_222 per_333
dm people ids --body '{"ids": ["per_111", "per_222"]}' --include-properties --property-limit 20
dm people ids per_111 per_222 --fields estimated_household_income,estimated_value
| Option | Description |
|---|---|
--include-properties |
Include associated properties |
--property-limit <n> |
Maximum associated properties to return per person, up to 100 |
--fields <csv> |
Comma-separated field IDs from dm fields |
dm people export
Export people as CSV (up to 1,000,000 records). Returns signed download URLs.
dm people export -f search.json --require-email
dm people export -f search.json --mobile-only --scrub-dnc --json
Contact filter options are the same as dm properties export.
Enrich Commands
All enrichment commands support three input modes: a positional argument for single-item lookup, --body/-f for JSON payloads, and -f with a .csv file for batch enrichment from CSV. Batches larger than 250 items are automatically chunked. Every enrichment command accepts --fields <csv> and sends the selected field IDs to the API. Email, phone, and name matches also include a free property_count; use --include-properties when you need the property records themselves.
dm enrich address [address]
Look up a property by street address.
# Single address
dm enrich address "123 Main St, Austin, TX 78704"
dm enrich address "123 Main St, Austin, TX 78704" --contact-audience none
dm enrich address "123 Main St, Austin, TX 78704" --fields estimated_value,equity
# Batch from JSON
dm enrich address --body '{"data": [{"full_address": "123 Main St, Austin, TX"}]}'
# Batch from CSV (auto-detected by .csv extension)
dm enrich address -f addresses.csv --contact-audience owners
# CSV columns: full_address (or street, city, state, zip)
| Option | Description |
|---|---|
--body <json> |
Request body as JSON string |
-f, --file <path> |
Read from JSON or CSV file |
--contact-audience <audience> |
owners, owners_and_family, renters, residents, none |
--fields <csv> |
Comma-separated field IDs from dm fields |
--json |
Output as JSON |
Use --contact-audience none whenever you only need the property. The response omits contacts and consumes zero people credits.
Pertanyaan tentang DealMachine MCP
Bagaimana cara menambahkan DealMachine MCP ke Claude Code?
Jalankan claude mcp add --transport http dealmachine-cli https://mcp.dealmachine.com/ di terminal Anda. Tambahkan --scope user agar tersedia di semua proyek.
Bagaimana cara menambahkan DealMachine MCP ke Cursor?
Buka ~/.cursor/mcp.json dan tambahkan server di bawah mcpServers, menggunakan config Cursor yang ditunjukkan di bagian instalasi di atas.
Apa yang bisa saya lakukan dengan DealMachine?
Anda dapat melakukan riset properti dan orang, pengayaan data, analisis perbandingan penjualan, serta mengelola daftar prospek dan webhook.
Bagaimana cara autentikasi dengan MCP server?
Server ini mendukung OAuth 2.1 untuk client seperti ChatGPT dan Claude, atau Anda dapat menggunakan API key DealMachine untuk client developer.
Bisakah saya mengestimasi biaya sebelum melakukan operasi berbayar?
Ya, server ini menyertakan skill yang sadar kredit (credit-aware) yang dapat menghitung jumlah kecocokan dan mengestimasi biaya sebelum mengeksekusi operasi berbayar dalam skala besar.