Name: Market Intelligence MCP
Author: BullrunData

LABSCO SUMMARY

The models sit behind Pro; the series they are built from do not.

recession_probability, market_regime, fed_stance, confirmation_status and housing_cycle are where BullrunData's own weighting lives, and on the free tier each answers 403 — worth knowing before an agent plans a workflow around dashboard_summary and finds placeholders in three of its blocks. Everything read from the published data still answers on 100 calls a day: yield curve, inflation, labour, housing, sector rotation, CFTC positioning and TIC flows. The Cascade Engine is the part with no obvious counterpart — it turns "what if oil hits $120" into a named chain of effects with confidence levels and precedents rather than a paragraph of commentary.

What it is

A macro and markets server over the BullrunData API. The hosted endpoint at https://bullrundata.com/api/mcp needs no install; the same tools run locally from npx -y @bullrundata/market-intelligence. Some calls read published economic series, others run BullrunData's own models, and the model-driven ones are Pro-tier.

What you get

dashboard_summary is the one-call macro snapshot; on the free tier the indicator block returns and the recession-probability, market-regime and Fed-stance blocks come back as an upgrade placeholder
recession_probability, market_regime, fed_stance, confirmation_status and housing_cycle are the proprietary models and answer HTTP 403 with an upgrade link below Pro — housing_cycle scores 0-100 across supply, demand, price, affordability and cost of capital
sectors_rotation reads cyclical against defensive S&P sectors and returns RISK_ON, RISK_OFF or NEUTRAL with leaders, laggards and 1w / 1m / 3m momentum; sectors_flows tags each sector ETF ACCUMULATION, DISTRIBUTION or NEUTRAL
institutional_cftc aggregates futures positioning; cftc_contracts_list then cftc_positioning_detail opens one contract's smart-money breakdown with week-over-week change; institutional_tic tracks net foreign purchases of US Treasuries
economic_indicator returns date-and-value pairs for any tracked series ID — VIXCLS, UNRATE, CPIAUCSL, T10Y2Y, MORTGAGE30US, BAMLH0A0HYM2 — with list_indicators, interest_rates, inflation_data, employment_data, housing_data, yield_curve and market_sentiment as ready-made cuts
The Cascade Engine maps a catalyst to its chain reaction across markets, regions and asset classes with confidence levels and historical precedents: cascade_list, cascade_search, cascade_by_category and cascade_analysis, over catalysts including Oil Supply Shock, Fed Emergency Rate Cut and Credit Market Freeze
investment_property_analysis and brrrr_analysis run locally and make no API call: cap rate, cash-on-cash ROI, NOI, DSCR and the 1% rule on one side, all-in cost, ARV margin, refinance cash-out, a 0-100 BRRRR score and the 70% rule on the other

Requirements

The hosted endpoint takes Authorization: Bearer <YOUR_BULLRUNDATA_API_KEY>, or negotiates OAuth in the browser on first connect; Claude Desktop reaches it through mcp-remote. The local package reads BULLRUNDATA_API_KEY instead. Free is 100 calls a day, Pro is 10,000 at $29/mo and Business is 100,000 at $99/mo; past the limit calls return HTTP 429 and the counter resets at 00:00 UTC. Elastic License 2.0 — use, modify and redistribute freely, but not as a hosted service of your own.

Setup effort

One command plus a key — claude mcp add --transport http bullrundata https://bullrundata.com/api/mcp --header "Authorization: Bearer <YOUR_BULLRUNDATA_API_KEY>"

See installation options

USE CASES

01

Getting a macro read before a positioning decision

One snapshot call for the headline block, then the curated indicator set with labels and units. On the free tier the model blocks are placeholders, so read the indicators directly.

dashboard_summaryrecession_indicatorslist_indicators

02

Checking whether the market is risk-on or risk-off right now

Cyclical against defensive sector performance gives the signal and its leaders and laggards; per-sector accumulation and distribution says whether flows agree; futures positioning says where institutional money leans.

sectors_rotationsectors_flowsinstitutional_cftc

03

Working out what a specific shock would set off

Search the catalyst library by keyword, then pull the full chain reaction with confidence levels and historical precedents across markets and asset classes.

cascade_searchcascade_analysiscascade_by_category

04

Pulling a time series without leaving the chat

Any tracked indicator by series ID over the range you ask for, returned as date-and-value pairs — VIXCLS for volatility, T10Y2Y for the spread, MORTGAGE30US for rates.

economic_indicatorlist_indicatorsyield_curve

05

Underwriting a rental deal at today's rates

Cap rate, DSCR, monthly cash flow and the 1% rule against the current 30-year mortgage rate, or the full BRRRR breakdown with the 70% rule if the plan is to refinance out.

investment_property_analysisbrrrr_analysisinterest_rates

06

Tracking one contract's positioning week to week

List the available futures contracts, then open a single one for the smart-money long and short breakdown and how it changed since last week.

cftc_contracts_listcftc_positioning_detailinstitutional_tic

27 tools

Counted once with only startup setup. Other versions, modes or optional credentials may expose a different list.

`brrrr_analysis`

Analyze a BRRRR (Buy, Rehab, Rent, Refinance, Repeat) real estate deal. Returns all-in cost, ARV margin, refinance cash-out, monthly cash flow, BRRRR Score (0-100), 70% rule check, and full breakdown. Pure local calculation — no API call.

BEFORE YOU INSTALL

What to know first

Tools you get

27 tools when we connected. housing_cycle is available but undocumented. The free tier limits calls to 100 a day. One troubleshooting section uses https://market.bullrundata.com/mcp instead of https://bullrundata.com/api/mcp.

Requires

A client that can reach a remote MCP server over HTTP
BULLRUNDATA_API_KEY in your environment

Where it runs

Of the 7 documented setups, 2 run the server on your machine and 5 connect to a hosted endpoint.

CONTEXT COST

Moderate Standing cost

Rated by the size of the 27 tool definitions that ride along on every turn, not by how many there are

INSTALLATION

Connect your client

Server setup

claude mcp add

Claude Code

TERMINAL

claude mcp add --transport http bullrundata https://bullrundata.com/api/mcp --header "Authorization: Bearer <YOUR_BULLRUNDATA_API_KEY>"

Hosted endpoint

ENDPOINT URL

https://bullrundata.com/api/mcp

Add this hosted endpoint to your MCP client using the credentials listed under Requirements.

MCP Clients

Cursor

CURSOR · JSON

{
  "mcpServers": {
    "bullrundata": {
      "url": "https://bullrundata.com/api/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_BULLRUNDATA_API_KEY>"
      }
    }
  }
}

VS Code

VS CODE · JSON

{
  "servers": {
    "bullrundata": {
      "type": "http",
      "url": "https://bullrundata.com/api/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_BULLRUNDATA_API_KEY>"
      }
    }
  }
}

Windsurf

WINDSURF · JSON

{
  "mcpServers": {
    "bullrundata": {
      "serverUrl": "https://bullrundata.com/api/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_BULLRUNDATA_API_KEY>"
      }
    }
  }
}

Claude Desktop — mcp-remote

CLAUDE DESKTOP — MCP-REMOTE · JSON

{
  "mcpServers": {
    "bullrundata": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://bullrundata.com/api/mcp",
        "--header",
        "Authorization: Bearer <YOUR_BULLRUNDATA_API_KEY>"
      ]
    }
  }
}

Claude Desktop — local npm

CLAUDE DESKTOP — LOCAL NPM · JSON

{
  "mcpServers": {
    "market-intelligence": {
      "command": "npx",
      "args": [
        "-y",
        "@bullrundata/market-intelligence"
      ],
      "env": {
        "BULLRUNDATA_API_KEY": "<YOUR_BULLRUNDATA_API_KEY>"
      }
    }
  }
}

Claude Code — local npm

CLAUDE CODE — LOCAL NPM · TERMINAL

claude mcp add market-intelligence -e BULLRUNDATA_API_KEY=<YOUR_BULLRUNDATA_API_KEY> -- npx -y @bullrundata/market-intelligence

Market Intelligence MCP

The models sit behind Pro; the series they are built from do not.

What it is

What you get

Requirements

Setup effort

Getting a macro read before a positioning decision

Checking whether the market is risk-on or risk-off right now

Working out what a specific shock would set off

Pulling a time series without leaving the chat

Underwriting a rental deal at today's rates

Tracking one contract's positioning week to week

27 tools

brrrr_analysis

cascade_analysis

cascade_by_category

cascade_list

cascade_search

cftc_contracts_list

cftc_positioning_detail

confirmation_status

dashboard_summary

economic_indicator

employment_data

fed_stance

housing_cycle

housing_data

inflation_data

institutional_cftc

institutional_tic

interest_rates

investment_property_analysis

list_indicators

market_regime

market_sentiment

recession_indicators

recession_probability

sectors_flows

sectors_rotation

yield_curve

What to know first

Tools you get

Requires

Where it runs

Moderate Standing cost

Connect your client

Server setup

claude mcp add

Claude Code

Hosted endpoint

MCP Clients

Cursor

VS Code

Windsurf

Claude Desktop — mcp-remote

Claude Desktop — local npm

Claude Code — local npm

Same category

Octagon

Market Sizing MCP Server

FeedOracle Macro Intelligence

Federal Reserve Economic Data

`brrrr_analysis`

`cascade_analysis`

`cascade_by_category`

`cascade_list`

`cascade_search`

`cftc_contracts_list`

`cftc_positioning_detail`

`confirmation_status`

`dashboard_summary`

`economic_indicator`

`employment_data`

`fed_stance`

`housing_cycle`

`housing_data`

`inflation_data`

`institutional_cftc`

`institutional_tic`

`interest_rates`

`investment_property_analysis`

`list_indicators`

`market_regime`

`market_sentiment`

`recession_indicators`

`recession_probability`

`sectors_flows`

`sectors_rotation`

`yield_curve`