Rekko Skill

Prediction market intelligence skill for AI agents. Deep causal research, arbitrage detection, screening, and strategy signals for Kalshi, Polymarket, and Robinhood.

What is Rekko?

In the OpenClaw ecosystem, three skills work together for autonomous prediction market trading:

Skill	Role	What it does
Rekko (this skill)	The Brain	Deep causal research, probability estimates, trading signals
PolyClaw	The Hands (Polymarket)	Executes trades on Polymarket via CTF minting + CLOB
Kalshi Trader	The Hands (Kalshi)	Executes trades on Kalshi

Rekko provides the why behind market movements. Execution skills handle the how.

Install

npx skills add Rekko-AI/rekko-skill

Or copy SKILL.md directly into your agent's skill directory.

Authentication

Two options:

API key (simple)

Get a key at rekko.ai, then set:

export REKKO_API_KEY=rk_free_your_key_here

x402 autopay (for autonomous agents)

Fund an EVM wallet with USDC on Base mainnet, then use the included Python client:

from eth_account import Account
from rekko_tools import RekkoClient

signer = Account.from_key("0x<your-private-key>")
async with RekkoClient(signer=signer) as client:
    markets = await client.list_markets()
    signal = await client.get_strategy("Will the Fed cut rates?")

Or set X402_PRIVATE_KEY in your environment for auto-detection.

Python client

Install dependencies:

pip install -r requirements.txt

The rekko_tools.py module provides RekkoClient — an async HTTP client with x402 payment support:

from rekko_tools import RekkoClient

async with RekkoClient() as client:
    # Browse markets ($0.01/call)
    markets = await client.list_markets(source="kalshi", limit=30)

    # Screen for high-value candidates ($0.10/call)
    screened = await client.screen_markets(platform="kalshi", min_score=50)

    # Deep analysis with strategy signal ($2.00/call)
    signal = await client.get_strategy("Will Bitcoin hit 100K by June?")

    # Cross-platform arbitrage scan ($5.00/call)
    arbs = await client.get_arbitrage(min_spread=0.03)

Example workflow

1. list_markets(source="kalshi", limit=30)
2. screen_markets(platform="kalshi", min_volume_24h=50000, min_score=50)
3. get_strategy(market_query="Will the Fed cut rates?")
4. IF recommendation == "BUY_YES" AND confidence > 0.5:
     -> chain to Kalshi Trader or PolyClaw for execution
5. place_shadow_trade(ticker, side, size_usd)  # track in Rekko portfolio

See SKILL.md for all 6 workflow patterns, complete endpoint reference, and response interpretation guide.

Links

• rekko.ai/docs — Full documentation
• OpenClaw integration guide — Detailed setup
• API Reference — Direct API access
• Discord — Community