Keel — Hyperliquid trading agent

This repository is the public mirror of thekeel-tradePython package: a singlepipx installgives you both a CLI and a stdio MCP server, so the same tools work from a terminal or from any MCP-capable agent.

Why agents create strategies, not trade them

Most agent-trading projects put an LLM in the execution loop. That makes systems slow, inconsistent, and hard to audit. Keel does the opposite:

You ──── compose ────► Strategy graph ──── compile ────► Deterministic artifact ▲ │ │ ▼ Agent edits Backtest engine via MCP tools (real HL data) │ ▼ Live execution (same artifact)

- Bit-for-bit parity between backtest and live.Same compiled artifact, same engine, same data path. There is no second implementation that can drift.
- Typed composition over freeform code.Strategies are graphs of versioned components. Compile errors catch bugs at author time instead of in production.
- Agents compose, the deterministic engine executes.Claude / Cursor / Codex help you build the strategy. They are not in the trade loop.

The MCPB bundle requires system Python 3.11+ (same prerequisite as the terminal install path below). First launch takes ~10-30 seconds while the bundle pip-installs runtime deps to~/.keel/mcpb-lib/py3.X/; subsequent launches are instant.

Terminal — pipx / uv (Claude Code, Codex, Cursor, Windsurf, etc.)

uv tool install keel-tradealso works. Python 3.11+.

Then register the stdio MCP command with your agent host:

# Claude Code claude mcp add keel -- keel mcp serve # Codex codex mcp add keel -- keel mcp serve

Agent:Callskeel_auth_login. Browser opens to app.usekeel.io, you click Allow, tokens land in~/.keel/config.yaml. Authenticated for 30 days with transparent refresh.

You:"Find me momentum signals for Hyperliquid top-30 perps and compose a backtest from 2024-08-15 to today."

Agent:Callskeel_components_search→keel_components_detail_batch→keel_strategy_compose→keel_backtest_run. Returns a share URL with the full tearsheet (equity curve, Sharpe, max drawdown, per-asset attribution).

The default toolset spans status, auth, components, strategy lifecycle, backtest, audit, accounts, sharing, and read-only live monitoring.Live-write tools(keel_live_deploy,keel_live_control) require an explicit opt-in toolset plus a local arming step — agents can't deploy your account without you authorizing it twice.

Every MCP outcome tool has a CLI mirror. Useful for terminals, SSH sessions, CI, scripts, or agents that prefer subprocess calls:

# Auth + status keel auth login keel status # Search components, compose, backtest keel components search "momentum" keel strategy compose --source-file my-strategy.py --dry-run keel backtest run str_abc123 --start-date 2024-08-15 --wait # Inspect a strategy keel strategy get str_abc123 keel strategy log str_abc123

Alpha. The CLI and MCP surface are stable and ship to PyPI on a regular cadence; the underlying engine and component library are actively developed.

Crypto data & trading MCP with 42+ tools: prices, DeFi, NFTs, Solana swaps

MCP server for Fuse Network: balances, tokens, staking, DeFi data, swaps and on-chain transactions.

Live crypto prices, Reddit sentiment, DeFi TVL, stock prices. Pay per call via x402 on Base.

MCP server for the Ophis intent-based DEX aggregator: quotes and gasless, MEV-protected swap orders across 11 chains.

Live scored Solana memecoin signals with safety profiles, conviction scoring, and paper trading for AI agents.

Connect any MCP-capable AI agent to live and historical Polymarket prediction-market orderbook data plus Hyperliquid perpetual-futures orderbooks (12 tools, 2 resources).

Manage Uniswap and Aerodrome liquidity positions with automated rebalancing and leverage.

Risk Intelligence MCP Server for crypto agents — 52 tools, 72B AI model, 560+ signals, derivatives, on-chain, autonomous trading

Cross-chain cryptocurrency swaps via Chainflip. Get quotes, execute swaps, and track progress.