Nephyr Backtest
- clm-studios/nephyr-backtest
io.github.clm-studios/nephyr-backtest is a MCP server that backtests prediction-market strategies on real on-chain Polymarket data.
About
Nephyr Backtest is a Model Context Protocol server that runs prediction-market strategy backtests against authentic on-chain Polymarket data. developers exploring event-contract products, arbitrage bots, or research newsletters can ask their agent to iterate strategy parameters and compare outcomes in simulation before touching real capital or complex indexing jobs. The package publishes to PyPI as nephyr-backtest with a uvx-friendly runtime hint and stdio transport, which fits quick validation in Claude Code or Cursor without a long-lived custom data pipeline. It complements generic finance MCPs that quote prices but do not replay historical market paths. Treat results as validation evidence for scope and pricing decisions, then move proven logic into your own backend in Build. Advanced users should still verify assumptions about fees, liquidity, and resolution rules outside the MCP.
- Backtests prediction market strategies using real on-chain Polymarket data
- Nephyr Backtest MCP server (PyPI nephyr-backtest, uvx runtime hint)
- stdio transport for local agent-driven experiment loops
- Version 0.1.1 from clm-studios/nephyr-backtest repository
- Separates research simulation from live wallet execution
Nephyr Backtest by the numbers
- Data as of Jul 7, 2026 (Skillselion catalog sync)
claude mcp add nephyr-backtest -- uvx nephyr-backtestAdd your badge
Show developers this MCP server is listed on Skillselion. Paste this into your README.
| Package | nephyr-backtest |
|---|---|
| Transport | STDIO |
| Auth | None |
| Repository | clm-studios/nephyr-backtest ↗ |
What it does
Backtest prediction-market strategies on real Polymarket on-chain data before you ship trading logic or dashboards.
Who is it for?
Best when you're validating Polymarket or prediction-market automation ideas and want MCP-driven experiment loops.
Skip if: Skip if you have no crypto or prediction-market interest, or teams needing live order execution rather than simulation.
What you get
After you wire the stdio server, your agent can run backtests on chain-backed data so you keep or kill strategies before production code.
- Historical strategy performance runs on Polymarket data
- Agent-assistable experiment iterations for go/no-go decisions
- Evidence pack before implementing live trading services
By the numbers
- Server version 0.1.1
- PyPI identifier nephyr-backtest with uvx runtime hint
- Uses real on-chain Polymarket data per server description
Recommended MCP Servers
$THREE TokenPrice, hold, and burn $THREE on Solana — the first MCP server whose actions burn a token.89
402 McpAI agents discover, pay for, and consume any Lightning-gated API autonomously
9192 Public ComputeRemote MCP for 9192 discovery, pricebook, payment rails, GET_PULSE quotes, and receipt checks.
Aarna McpAarna - 34 tools for swap quotes, liquidity pools, and trading dataAave McpAave - 6 tools for lending rates, supply, and borrow data
AbraFlexiMCP server for AbraFlexi ERP — invoices, contacts, products, bank transactions.1How it compares
Polymarket backtest MCP, not China A-share quant data or federal grants research.
FAQ
Who is io.github.clm-studios/nephyr-backtest for?
Developers and quant hobbyists testing event-market strategies who use MCP agents and want Polymarket history without building an indexer first.
When should I use io.github.clm-studios/nephyr-backtest?
Use it in validate and prototype when you need evidence that a strategy behaved sensibly on real on-chain Polymarket data before you implement trading infra.
How do I add io.github.clm-studios/nephyr-backtest to my agent?
Install nephyr-backtest from PyPI (uvx runtime hint), register the stdio MCP server in your host, and invoke backtest tools from your agent per the repository README.