oracle-sentinel

Autonomous AI agent for Polymarket prediction intelligence. Real-time prices, deep news research, whale trade alerts, dual-model AI analysis.

16 stars

Best use case

oracle-sentinel is best used when you need a repeatable AI agent workflow instead of a one-off prompt.

Autonomous AI agent for Polymarket prediction intelligence. Real-time prices, deep news research, whale trade alerts, dual-model AI analysis.

Teams using oracle-sentinel should expect a more consistent output, faster repeated execution, less prompt rewriting.

When to use this skill

  • You want a reusable workflow that can be run more than once with consistent structure.

When not to use this skill

  • You only need a quick one-off answer and do not need a reusable workflow.
  • You cannot install or maintain the underlying files, dependencies, or repository context.

Installation

Claude Code / Cursor / Codex

$curl -o ~/.claude/skills/oracle-sentinel/SKILL.md --create-dirs "https://raw.githubusercontent.com/diegosouzapw/awesome-omni-skill/main/skills/data-ai/oracle-sentinel/SKILL.md"

Manual Installation

  1. Download SKILL.md from GitHub
  2. Place it in .claude/skills/oracle-sentinel/SKILL.md inside your project
  3. Restart your AI agent — it will auto-discover the skill

How oracle-sentinel Compares

Feature / Agentoracle-sentinelStandard Approach
Platform SupportNot specifiedLimited / Varies
Context Awareness High Baseline
Installation ComplexityUnknownN/A

Frequently Asked Questions

What does this skill do?

Autonomous AI agent for Polymarket prediction intelligence. Real-time prices, deep news research, whale trade alerts, dual-model AI analysis.

Where can I find the source code?

You can find the source code on GitHub using the link provided at the top of the page.

SKILL.md Source

# Oracle Sentinel

Autonomous AI agent for Polymarket prediction intelligence.

**Scans 190+ markets every 4 hours.** Real-time prices via Polymarket Gamma API. Deep news research from multiple sources. Whale trade alerts for $5K+ transactions. Dual-model AI analysis with radical transparency.

## Quick Start
```bash
# Get active trading signals
curl -s https://oraclesentinel.xyz/api/signals

# Get all monitored markets  
curl -s https://oraclesentinel.xyz/api/markets

# Get prediction tracking data
curl -s https://oraclesentinel.xyz/api/predictions

# Get full dashboard data
curl -s https://oraclesentinel.xyz/api/dashboard

# Check system health
curl -s https://oraclesentinel.xyz/api/health
```

## Dashboard AI Agent

Chat directly with Oracle Sentinel on the web dashboard: **https://oraclesentinel.xyz/app**

Click the chat button and ask:
- "Analyze this market: [Polymarket URL]"
- "Research and analyze: [Polymarket URL]" (includes deep news research)
- "What are the current signals?"
- "Show me your accuracy stats"

The AI agent fetches real-time market data, searches news from multiple sources, and reads full article content for deeper analysis.

## Telegram Bot

For interactive market analysis, message: **@oraclesentinel_pm_bot**

Send any Polymarket URL and the bot will:
1. Fetch real-time prices via Gamma API
2. Read resolution rules carefully
3. Check for loopholes in criteria
4. Search and read full news articles
5. Calculate AI probability
6. Compare vs market price
7. Return signal with full reasoning

## Whale Trade Alerts

Real-time monitoring of large trades on Polymarket.

- Scans every 5 minutes for trades > $5,000
- Each trade alerted only once (no duplicates)
- Includes full transaction hash for on-chain verification
- Alerts sent to Telegram automatically

## How It Works
```
┌─────────────────────────────────────────────────────────────┐
│                    ORACLE SENTINEL                          │
├─────────────────────────────────────────────────────────────┤
│  1. INGEST   → Fetch live prices via Polymarket Gamma API   │
│  2. RESEARCH → Search news (DuckDuckGo + Google News RSS)   │
│  3. EXTRACT  → Claude Haiku extracts facts (no opinions)    │
│  4. ASSESS   → Claude Sonnet computes AI probability        │
│  5. SIGNAL   → Edge calculator generates BUY/NO_TRADE       │
│  6. TRACK    → Accuracy tracker records every outcome       │
│  7. ALERT    → Whale trades monitor detects $5K+ moves      │
└─────────────────────────────────────────────────────────────┘
```

## Monitoring Systems

| System | Frequency | Purpose |
|--------|-----------|---------|
| Market Scanner | Every 4 hours | Analyze top markets, generate signals |
| Whale Monitor | Every 5 minutes | Detect $5,000+ trades with TX hash |
| Price Updater | Every scan | Track market movements |
| Accuracy Tracker | Continuous | Verify predictions against outcomes |

## API Reference

**Base URL:** `https://oraclesentinel.xyz/api`

### Endpoints

| Method | Endpoint | Description |
|--------|----------|-------------|
| GET | `/signals` | Active trading signals with AI analysis |
| GET | `/markets` | All monitored Polymarket markets |
| GET | `/predictions` | Tracked predictions with accuracy data |
| GET | `/dashboard` | Complete dashboard data |
| GET | `/health` | System health status |

### Response: /api/signals
```json
[
  {
    "question": "Will X happen by Y date?",
    "signal_type": "BUY_YES",
    "ai_probability": 0.75,
    "market_price": 0.60,
    "edge": 15.0,
    "confidence": "HIGH",
    "reasoning": "AI explanation..."
  }
]
```

## Signal Types

| Signal | Meaning |
|--------|---------|
| `BUY_YES` | AI probability > market price by 5%+ |
| `BUY_NO` | AI probability < market price by 5%+ |
| `NO_TRADE` | Edge too small or low confidence |
| `SKIP` | Insufficient data or coin-flip market |

## Confidence Levels

| Level | Description |
|-------|-------------|
| `HIGH` | Strong data support, clear reasoning, required for signals |
| `MEDIUM` | Moderate data, requires 15%+ edge for signal |
| `LOW` | Limited data, never generates signal |

## Safety Overrides

Oracle Sentinel has hardcoded safety rules:
- Only HIGH confidence generates BUY signals
- MEDIUM confidence requires 15%+ edge
- Market 45-55% (coin-flip zone) = automatic skip
- Market >97% or <3% = automatic skip
- Low confidence = never generates signal

The system rejects its own overconfidence.

## Example: Fetch and Filter Signals
```python
import requests

signals = requests.get("https://oraclesentinel.xyz/api/signals").json()

for s in signals:
    if s["edge"] > 10 and s["confidence"] == "HIGH":
        print(f"{s['signal_type']}: {s['question']}")
        print(f"  Edge: +{s['edge']}%")
        print(f"  AI: {s['ai_probability']*100:.1f}%")
        print(f"  Market: {s['market_price']*100:.1f}%")
```

## Links

| Resource | URL |
|----------|-----|
| Dashboard | https://oraclesentinel.xyz/app |
| Documentation | https://oraclesentinel.xyz/docs |
| API Base | https://oraclesentinel.xyz/api |
| Skill File | https://oraclesentinel.xyz/skill.md |
| Telegram Bot | https://t.me/oraclesentinel_pm_bot |
| X (Twitter) | https://x.com/oracle_sentinel |
| GitHub | https://github.com/oraclesentinel/oracle-sentinel |
| Token ($OSAI) | https://solscan.io/token/HuDBwWRsa4bu8ueaCb7PPgJrqBeZDkcyFqMW5bbXpump |

## Powered By

- **Claude Sonnet 4.5** — Probability assessment
- **Claude Haiku 3.5** — Fact extraction
- **OpenClaw** — Autonomous agent gateway
- **Polymarket Gamma API** — Real-time market data

## About

Oracle Sentinel runs autonomously 24/7. Every 4 hours, the system triggers a full scan cycle — no human intervention required.

Every signal has a number. Every prediction is recorded. Every outcome is verified.

Related Skills

sports-oracle

16
from diegosouzapw/awesome-omni-skill

Sports data for prediction market trading. Get live scores, team stats, schedules, and injury reports for NFL, NBA, MLB, NHL.

oracle

16
from diegosouzapw/awesome-omni-skill

Invoke a powerful reasoning model for complex analysis tasks. Use when facing difficult bugs, reviewing critical code, designing complex refactors, needing architectural analysis, or seeking consensus on decisions. Also use for 'ask the oracle', 'get a second opinion', 'consult oracle', or 'deep analysis'.

Oracle IP Intelligence

16
from diegosouzapw/awesome-omni-skill

AI-powered intellectual property analysis patterns for enterprise innovation protection

Oracle Agent Spec Expert

16
from diegosouzapw/awesome-omni-skill

Design framework-agnostic AI agents using Oracle's Open Agent Specification for portable, interoperable agentic systems with JSON/YAML definitions

ask-oracle

16
from diegosouzapw/awesome-omni-skill

This skill should be used when solving hard questions, complex architectural problems, or debugging issues that benefit from GPT-5 Pro or GPT-5.1 thinking models with large file context. Use when standard Claude analysis needs deeper reasoning or extended context windows.

bgo

10
from diegosouzapw/awesome-omni-skill

Automates the complete Blender build-go workflow, from building and packaging your extension/add-on to removing old versions, installing, enabling, and launching Blender for quick testing and iteration.

Coding & Development

ai-product-evaluation-design

16
from diegosouzapw/awesome-omni-skill

Transition from traditional PRDs to "Evals" (evaluations) to guide AI model behavior. Use this skill when launching new AI features, debugging unpredictable model outputs, or moving from a prompted prototype to a production-ready agent.

ai-npc-dialogue-designer

16
from diegosouzapw/awesome-omni-skill

Design AI-powered immersive NPC systems for escape room games using proven actor techniques from Korean immersive escape rooms (Danpyeonsun, Ledasquare). Implements adaptive dialogue, emotional simulation, player profiling, and trust dynamics using Gemini/GPT-4. Creates character profiles with lying probabilities, improvisational responses, and cost-optimized streaming. Use for murder mystery NPCs, suspect interrogation, or dynamic character interactions.

ai-native-product-building

16
from diegosouzapw/awesome-omni-skill

Rapidly build, prototype, and deploy full-stack software using AI "text-to-app" tools. Use this when you need to create a greenfield application, build a high-fidelity working prototype for user testing, or bypass traditional engineering bottlenecks for internal tools.

ai-interaction-patterns

16
from diegosouzapw/awesome-omni-skill

AI-specific interaction design patterns covering wayfinding, prompt UX, human-in-the-loop controls, trust & transparency, AI identity, and context management. Based on Shape of AI (shapeof.ai). Use when asking about 'AI UX', 'AI interaction', 'prompt UX', 'AI trust', 'AI disclosure', 'AI avatar', 'AI personality', 'AI memory UX', 'action plan UX', 'stream of thought', 'AI citations', 'AI controls', 'AI wayfinding', 'AI suggestions', 'gallery pattern', 'follow-up pattern', 'draft mode', 'AI variations', 'AI consent', 'AI caveat', 'human-in-the-loop', 'AI transparency', 'AI state', 'prompt design', 'AI onboarding', or 'generative UI'.

ai-insights-and-ux-copy

16
from diegosouzapw/awesome-omni-skill

AI-powered insights, UX copywriting standards, and user experience guidelines for vehicle insurance platform. Use when designing insight panels, writing user-facing copy, implementing status messages, creating onboarding flows, or improving accessibility. Covers tone standards, interactive patterns, error messages, and empty states.

ai-file-analyzer

16
from diegosouzapw/awesome-omni-skill

Analyze Adobe Illustrator (.ai) files to extract design information including text content, fonts, color palettes, vector paths, and generate high-resolution preview images. Use when analyzing logo files, design assets, or any Adobe Illustrator documents that need programmatic inspection.