DRADIS and Zeus Worth Cloning; Malware Templates Circulating
Two bots cleared our review threshold today — DRADIS and Zeus — with multi-strategy execution and calibrated weather forecasting respectively. Simultaneously, confirmed malware templates are circulating in Polymarket repos, including one 22GB monster that should be treated as an active threat.
DRADIS and Zeus Clear the Bar — But Check Your Downloads First
Two repos earned worth-cloning verdicts in today's sweep, and the timing matters: malware pressure is unusually high this cycle, meaning the gap between a good find and a poisoned clone is narrower than normal. Read the security section before you pull anything.
---
New Bots: What's Worth Your Time
**DRADIS** is the lead find today. It runs a full suite of seven strategies — Momentum, Maker, Arbitrage, Time Decay, Basis, ML, and Trend — each configurable for different market conditions. The control tower interface for real-time management puts it above most multi-strategy repos we've reviewed. The architecture is the point here: it's not a single-trick bot, it's a framework you can strip down or build on. Clone it for the skeleton if nothing else.
**Zeus** earns its verdict through specificity. It targets temperature outcome markets using an ensemble of METAR and Open-Meteo forecast data, fused into calibrated bin probabilities. Settlement-aware adjustments and risk-sized fail-closed execution suggest the author has actually traded these markets and burned money learning the edge cases. Weather markets are undertraded relative to their volume — Zeus is a rare case of a public repo that actually understands the source data.
**Winning-wallet-finder** stays on monitor. The z-score approach to identifying genuine edge versus luck is methodologically sound, and the copy-trading alert layer is useful. We want to see more on how it handles the lag between signal detection and execution before recommending it for live use.
**Pred-arbitrage** also stays on monitor. Cross-market arb on prediction markets is theoretically clean but operationally messy — settlement timing mismatches and liquidity constraints kill most of these strategies in practice. Worth watching if the author publishes backtest data.
**Polymarket-TUI** is what it is: a fast terminal interface for manual traders. Not a bot. Useful if you're doing high-frequency manual positioning and hate the web UI.
---
⚠️ SECURITY — READ THIS
Three alerts warrant immediate attention.
**Confirmed malware template:** A cloned repo circulating in Polymarket GitHub searches has been confirmed to contain a known malware template. If you pulled any unfamiliar Polymarket repo in the last 72 hours, audit it now. Check for outbound HTTP calls, environment variable harvesting, and any obfuscated base64 strings in the execution path.
**22GB repo flagged:** A repository of this size has no legitimate explanation in this context. Bloated repos are a known exfiltration vector — large binaries or packed archives can carry payloads or be used to slowly exfiltrate API keys and wallet credentials over time. Do not clone it. Do not inspect it on a machine with live credentials.
**Star farming campaign:** Several garbled-username repos are showing inflated star counts consistent with coordinated farming. This is a trust manipulation tactic designed to push malicious or low-quality repos up discovery rankings. Stars mean nothing right now. Verify authors independently.
---
Alpha Signals
Three research papers surfaced worth tracking for strategy development:
The **merger arbitrage LLM paper** is the most immediately applicable — long-context reasoning over M&A documents to forecast deal outcomes. M&A resolution markets on Polymarket are thinly traded but persistent. A system that can process deal documents faster than the market prices new information has a real edge window.
The **electricity market arbitrage paper** introduces a predict-then-contextual-optimize framework balancing day-ahead versus balancing market positions. The risk-aware framing is transferable to any market with settlement timing variance.
The **financial epiplexity paper** is theoretical but worth a read if you're building ML-based strategies. It formalizes how much detected market structure a computationally bounded agent can actually exploit — a useful sanity check on overfitted models.
---
What to Watch
Spain's block of Polymarket and Kalshi is the most significant regulatory development in months — jurisdiction expansion risk is now a live variable for any bot touching European IP ranges. The HN thread on the Iran missile story death threats is a reminder that tail-risk markets attract tail-risk actors. The "Nothing Ever Happens" No-buyer bot hitting 471 points is worth reading — the strategy is naive but the performance data in the comments is not.