# foresea MCP server

Probability forecasts with evidence, prediction-market edge, and a live track record.

## Links
- Registry page: https://www.getdrio.com/mcp/io-github-pareelamre-foresea
- Repository: https://github.com/pareelamre/analyzing-llm-rationale

## Install
- Endpoint: https://foresea.ink/mcp/
- Auth: Not captured

## Setup notes
- Remote endpoint: https://foresea.ink/mcp/

## Tools
- foresea_forecast - Forecast one resolvable question and return probability, rationale, evidence, and optional market edge. Endpoint: https://foresea.ink/mcp/
- foresea_analyze_market - Run Foresea's end-to-end market agent: resolve market, gather evidence, forecast, compute edge, and recommend. Endpoint: https://foresea.ink/mcp/
- foresea_scan_markets - Scan Polymarket, Kalshi, or both venues for live markets where Foresea sees a large edge. Endpoint: https://foresea.ink/mcp/
- foresea_track_record - Return Foresea's public track record and calibration summary. Endpoint: https://foresea.ink/mcp/
- foresea_edge_board - Live model-vs-market edge board: open markets ranked by how much Foresea's
        fair probability disagrees with the price, each tagged with the resolved
        track record of disagreements that size (skill-vs-market + significance),
        plus by-edge calibration and market lead/lag — i.e. where the disagreement
        is, and whether disagreement that size has actually paid. Endpoint: https://foresea.ink/mcp/

## Resources
- foresea://track-record - Foresea's public resolved-forecast track record. MIME type: application/json
- foresea://openapi.json - Foresea's public OpenAPI schema. MIME type: application/json

## Prompts
- foresea_forecast_prompt - Create a compact prompt that asks an agent to use Foresea for a forecast. Arguments: question

## Metadata
- Owner: io.github.pareelamre
- Version: 1.27.2
- Runtime: Streamable Http
- Transports: HTTP
- License: Not captured
- Language: Not captured
- Stars: Not captured
- Updated: Jun 6, 2026
- Source: https://registry.modelcontextprotocol.io