Cross-Market Arb Tools Surge; Weather Bot Niche Matures
Today's GitHub sweep surfaces a notable cluster of cross-platform arbitrage scanners targeting Polymarket, Kalshi, and PredictIt simultaneously, signaling that serious operators are treating these venues as a unified liquidity pool rather than isolated markets. Separately, two independent weather-market bots with documented methodology indicate the temperature-market niche has moved past proof-of-concept into systematic exploitation.
GitHub Intelligence — 2026-06-20
Lead Story: Cross-Market Arbitrage Infrastructure Is Maturing
Four repositories pushed or updated today target cross-venue arbitrage between Polymarket, Kalshi, and PredictIt: `matthewnyc2/arbitrage`, `Steeve-Crypto/Quant-CLI`, `pjmerica/pred-arbitrage`, and `Sourav19o7/polymarket-kalshi-arb`. This is not a coincidence of timing. The clustering suggests a shared awareness — likely driven by community discussion — that NegRisk structure on Polymarket and Kalshi's regulated US framework create persistent mispricings on correlated binary outcomes, particularly in political and macro markets.
`matthewnyc2/arbitrage` is the most operationally credible of the group: it claims 60 passing tests, a live browser demo, and production-grade asyncio scanning specifically targeting NegRisk multi-outcome violations. NegRisk arb is non-trivial — it requires monitoring the full complement-set of outcomes and accounting for Polymarket's fee structure before a spread becomes real. The fact that someone has built a test suite around this implies live edge-testing, not academic exploration.
`nayrbryanGaming/arbiter` takes a different angle, focusing exclusively on political markets and cross-market no-arbitrage violations between related election markets. Its fractional Kelly sizing and mandatory backtested edge requirement suggest an operator trying to solve the position-sizing problem correctly, not just find the spread. Paper-trading mode default is noted — treat this as developmental, not deployed.
**Operator implication:** If even partially functional, these scanners will compress NegRisk spreads and Polymarket-Kalshi parity gaps faster than manual monitoring can capture them. Any edge you currently hold in those structures should be treated as having a shorter half-life.
---
Weather Markets: A Niche With Real Methodology
Two weather-focused repos stand out. `RiekertQuant/polymarket-weather-bot-poc` has accumulated 21 stars and 5 forks since January — meaningful traction by Polymarket tooling standards. `BallesJr/polymarket-weather-edge` documents a specific approach: betting the NO side of overpriced temperature markets using live METAR observations and Open-Meteo forecasts fed into a Gaussian probability model.
The methodology is sound in outline. METAR data is real-time, machine-readable, and free. Open-Meteo provides calibrated ensemble forecasts. A Gaussian model over a temperature threshold is a reasonable first approximation. The critical unknown is how well Polymarket's crowd prices temperature markets — if the market is already incorporating NWS forecasts, the edge collapses. The fact that `BallesJr` is running paper trading rather than live capital suggests they haven't confirmed a persistent edge yet.
**Operator implication:** Weather markets may be soft. If you hold positions in temperature or precipitation markets, be aware that at least two teams are running systematic models against them. Verify your pricing against Open-Meteo before entry.
---
PMXT: The Infrastructure Play to Watch
`pmxt-dev/pmxt` continues to be the most significant repository in the Polymarket tooling ecosystem, now at 1,904 stars and 225 forks. Framed as a CCXT equivalent for prediction markets — a unified API across Polymarket, Kalshi, and others — this is the abstraction layer the cross-market arb tools above are likely building on. Watch its commit history for new venue integrations or breaking API changes, as downstream bot behavior will follow.
---
Security Note
`quipmnxailcrrgky/tradingbot` warrants scrutiny. A 98-star, 60-fork repo from an account with a randomized-looking username (`quipmnxailcrrgky`), created in March and sized at 34MB, matches the profile of credential-harvesting or wallet-draining packages that have historically circulated in trading-bot communities. Do not execute code from this repository without full source review. The star count is not sufficient vetting.
---
Summary Assessment
Today's feed reflects an ecosystem that is professionalizing around specific, exploitable structures: NegRisk arbitrage, cross-venue parity gaps, and niche weather markets. The tooling is improving faster than most manual operators will track. The actionable priority is to audit any positions you hold in markets with cross-venue equivalents and to treat weather market pricing as increasingly contested.