Meridian Edge — Prediction Market Consensus

io.github.tmbot12Published by GitHub account tmbot12v0.1.0Updated Oct 24, 2026

Real-time prediction market consensus data for AI agents

The publisher's own server: Edge Network

VerifiedSTDIODesktop onlyFinanceData & Analytics

Overview

AI-generated overview

Local MCP server that gives an assistant live prediction-market consensus data on sports, politics and economics events from Meridian Edge.

What you can do
  • Read consensus probabilities for sports, politics and economics events
  • Find events where prediction markets diverge, plus recent directional market moves
  • List active tracked markets and recently settled events with verified outcomes
  • Return spread, confidence and trend details alongside the consensus figures
When to use it
Useful when you want an assistant to answer questions about what prediction markets currently imply for a game, election or macro event, or to surface disagreements and recent moves across markets. Not needed for general web research or betting execution.
Requirements
Runs locally as a Python package (meridian-edge-mcp) via uvx, pip install, or from source; a desktop MCP client such as Claude Desktop, Cursor or Windsurf. An API key in MERIDIAN_EDGE_API_KEY is optional at first: a limited free demo key works out of the box, and a free key gives 100 calls/day. Network access to the Meridian Edge service.
Before you install
The API key in MERIDIAN_EDGE_API_KEY is a secret stored in your client config. Higher tiers are paid subscriptions ($29 to $499 per month) with daily call limits, so usage can hit caps or incur cost. Data is aggregated from third-party markets and the README states it is for informational purposes only, not investment advice.

Installation

In SourceWeft

  1. Open Meridian Edge — Prediction Market Consensus in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.

Other MCP clients

Follow the launch instructions in the repository.

README

[PyPI] [License: MIT] [MCP]

Meridian Edge MCP Server

Query real-time prediction market consensus data directly from Claude and other AI assistants via the Model Context Protocol.

Ask Claude: "What does the prediction market say about the Lakers game tonight?" and get a live answer.


What It Does

This MCP server gives Claude access to Meridian Edge — aggregated prediction market consensus data from multiple regulated prediction markets, updated every 10 minutes.

5 tools available to Claude:

ToolWhat it returns
get_consensusAggregated consensus probabilities for sports/politics events
get_opportunitiesEvents where prediction markets show notable divergence
get_signalsRecent directional market moves
get_marketsActive markets currently being tracked
get_settlementsRecently settled events with verified outcomes

Quick Start

Option 1 — uvx (recommended, no install)

json
{  "mcpServers": {    "meridian-edge": {      "command": "uvx",      "args": ["meridian-edge-mcp"]    }  }}

Option 2 — pip install

bash
pip install meridian-edge-mcp
json
{  "mcpServers": {    "meridian-edge": {      "command": "meridian-edge-mcp"    }  }}

Option 3 — run from source

bash
git clone https://github.com/meridian-edge/meridian-edge-mcpcd meridian-edge-mcppip install -e .
json
{  "mcpServers": {    "meridian-edge": {      "command": "python",      "args": ["-m", "meridian_edge_mcp.server"]    }  }}

Configure Claude Desktop

Open your Claude Desktop config file:

  • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
  • Windows: %APPDATA%\Claude\claude_desktop_config.json

Add the meridian-edge entry to mcpServers:

json
{  "mcpServers": {    "meridian-edge": {      "command": "uvx",      "args": ["meridian-edge-mcp"],      "env": {        "MERIDIAN_EDGE_API_KEY": "your_api_key_here"      }    }  }}

Restart Claude Desktop. You'll see the Meridian Edge tools listed in Claude's tool panel.

Cursor IDE

Add to .cursor/mcp.json:

json
{  "meridian-edge": {    "command": "uvx",    "args": ["meridian-edge-mcp"]  }}

Windsurf IDE

Add to Windsurf MCP config:

json
{  "meridian-edge": {    "command": "uvx",    "args": ["meridian-edge-mcp"]  }}

Try It Now

Ask Claude:

  • "What's the prediction market consensus on the Lakers game?"
  • "Show me events where prediction markets disagree"
  • "What markets settled today?"

Example response:

Lakers vs Celtics  Consensus: 62% (YES)  Sources: 5 regulated markets  Confidence: HIGH  Spread: 8.3%

API Key

The server works out of the box with a free demo key (limited data).

For full access, get a free API key at meridianedge.io — no credit card required, instant signup.

Set it as an environment variable:

bash
export MERIDIAN_EDGE_API_KEY=your_key_here

Or in your Claude Desktop config (see above).

Free tier: 100 calls/day — enough for personal use.


Example Prompts

Once configured, ask Claude:

Game consensus:

"What's the prediction market consensus on tonight's NBA games?"

"What are prediction markets saying about the Lakers vs Warriors?"

"Show me NFL prediction market probabilities for this week"

Divergence / opportunities:

"Which prediction markets are showing the most disagreement right now?"

"Show me events where prediction markets disagree — NBA only"

"What are today's highest-scoring divergence opportunities?"

Signals:

"What prediction market signals fired in the last hour?"

"Show me the most recent market moves"

Markets:

"What prediction markets are active right now?"

"List active NHL prediction markets"

History:

"How did recent prediction market consensus perform? Show settled events"

"What outcomes were predicted correctly in the last batch?"


Example Output

PREDICTION MARKET CONSENSUS — NBA (3 events)
1. NBA: ATL vs DET (2026-03-25)   Consensus: 46.6% YES  |  Spread: 1.6%  |  Confidence: MEDIUM   Trend (30min): ↓ -23.3%  |  Platforms: 2  |  ▃▃▄▄▇▆   Updated: 2026-03-26 01:15 UTC
2. NBA: BOS vs OKC (2026-03-25)   Consensus: 39.0% YES  |  Spread: 1.2%  |  Confidence: LOW   Trend (30min): ↑ +3.5%  |  Platforms: 2  |  ▆▆▆▆▆▄   Updated: 2026-03-26 01:15 UTC
3. NBA: CLE vs MIA (2026-03-25)   Consensus: 57.0% YES  |  Spread: 0.8%  |  Confidence: HIGH   Trend (30min): → stable  |  Platforms: 2  |  ▆▆▆▇▇▇   Updated: 2026-03-26 01:15 UTC
For informational purposes only. Not investment advice.Source: Meridian Edge — meridianedge.io

Data Coverage

  • Sports: NBA, NFL, MLB, NHL, MLS, college sports, boxing
  • Politics: US elections, ballot measures
  • Economics: Federal Reserve rate decisions, macro indicator markets
  • Update frequency: Every 10 minutes during active market hours

Pricing

TierPriceCalls/dayFeatures
Free$0100Consensus probabilities, active markets
Starter$29/mo500+ Opportunities, signals, spread data
Pro$99/mo5,000+ Fair value, platform breakdown, history
Teams$499/mo50,000+ Team seats, priority support

Full pricing →


AI Platform Integrations

Use Meridian Edge consensus data directly inside major AI platforms — no code required:

PlatformLinkNotes
ChatGPTOpen Custom GPTNo setup — just open and ask
Claude (this repo)MCP install guideClaude Desktop / Cursor via MCP
GeminiOpen GemNo setup — just open and ask

Links


Also Available On


License

MIT — see LICENSE

For informational purposes only. Not investment advice. Data aggregated from publicly available prediction market sources. © 2026 VeraTenet LLC d/b/a Meridian Edge.

Source: README.md at commit b955d70

Tools

0
SourceWeft has not started this package in its sandbox yet, so its tools are not known.

Version history

1
  1. v0.1.0LatestSep 16, 2026

Score

50% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 days25 pts

    Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.

  • Auth declared10 pts

    Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.

  • Tools listed15 pts

    Answer tools/list without a credential, or ship a package that starts in a sandbox without one.

  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts

Other Edge integrations