What happens if the AI analysis fails?

Xpulse uses a 'fail-closed' design, meaning if the AI encounters an error or timeout, it will remain silent rather than sending unverified or noisy alerts.

How does Xpulse scan X without API costs?

It uses DuckDuckGo search queries restricted to the x.com domain to find recent posts, bypassing the need for expensive official Twitter API subscriptions.

What local infrastructure is required?

You must have Ollama installed and running with the Qwen model pulled to handle the signal analysis and materiality gating stages locally.

Can I use this for markets other than Kalshi?

While the position-matching stage is currently optimized for Kalshi, the signal detection and materiality gates can be used for general market monitoring via the signal cache.

Does Xpulse share my Kalshi positions?

No. The skill runs locally and uses your Kalshi API credentials to fetch positions for matching purposes only; all AI analysis happens on your local machine via Ollama.

Xpulse Prediction Market Scanner

Name: Xpulse Prediction Market Scanner
Author: dvcrn

bydvcrn

•

Analytics & Monitoring

Monitors real-time social signals on X/Twitter and filters them against active Kalshi positions using local AI to deliver high-signal trading alerts.

Xpulse is a specialized Claude Code skill designed for prediction market traders who need real-time situational awareness without the noise of generic social media feeds. It utilizes a sophisticated three-stage pipeline that scans X via DuckDuckGo, performs materiality gating with local Qwen AI to prevent alert fatigue, and matches signals directly to your active Kalshi positions. By leveraging local LLMs and free search queries, it provides a cost-effective, privacy-focused solution for tracking market-moving events across topics like tariffs, inflation, and geopolitical policy changes.

Key Features

01Fail-closed design logic to ensure silence over notification noise

02Local AI analysis using Ollama and Qwen for zero-cost materiality gating

03Three-stage intelligence pipeline for high-fidelity signal detection

043 GitHub stars

05Position-aware filtering that only alerts on active Kalshi trades

06Historical signal caching for integration with morning market briefs

Use Cases

01Real-time monitoring of policy-driven market shifts on Kalshi

02Cost-free social signal scanning without expensive X/Twitter API fees

03Automated deduplication of social media news to prevent alert fatigue

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add dvcrn/openclaw-skills-marketplace xpulse

For use in Claude.ai and ChatGPT

Key Features

01Fail-closed design logic to ensure silence over notification noise

02Local AI analysis using Ollama and Qwen for zero-cost materiality gating

03Three-stage intelligence pipeline for high-fidelity signal detection

043 GitHub stars

05Position-aware filtering that only alerts on active Kalshi trades

06Historical signal caching for integration with morning market briefs

Use Cases

01Real-time monitoring of policy-driven market shifts on Kalshi

02Cost-free social signal scanning without expensive X/Twitter API fees

03Automated deduplication of social media news to prevent alert fatigue

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add dvcrn/openclaw-skills-marketplace xpulse

For use in Claude.ai and ChatGPT