Sentinel Skill banner
acevod acevod

Sentinel Skill

AI community

Description

Risk-aware AI trading copilot for Binance Agent OS — analyzes your portfolio, challenges risky trades before execution, tracks market conditions, journals every trade, and learns from your performance. Built for Binance's Agent OS Mini Hackathon.

Installation

This entry records only its repository, not the path inside it, so there is no exact command to give. Open the source below and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

README

Sentinel

![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg) ![Claude Skill](https://img.shields.io/badge/Claude-Skill-blueviolet) ![MCP](https://img.shields.io/badge/MCP-Binance%20Agent%20OS-F0B90B)

A Claude skill that turns Claude into a full trading copilot for [Binance Agent OS](https://binance.com/agent-os) (Binance's MCP server for AI applications) — not just an order executor, but a system that watches your portfolio, challenges risky trades, reads market conditions, remembers every trade, and learns from your track record over time.

Built for Binance's Agent OS Mini Hackathon (Track A).

Why "Sentinel"

The skill started as a single risk-check layer for trade execution. It grew into five modules covering the full loop a trader actually goes through: check state → get context → decide → act → record → learn. "Sentinel" is the umbrella name for that whole system.

The five modules

Module Triggers on What it does
Portfolio Analysis "what's my balance", "how's my portfolio" Full breakdown of holdings, exposure, P&L, and risk flags — never just a raw number
Trade Guardian "buy $X of Y", "open a 5x long", "close my position" Risk-aware execution: checks market + portfolio context, plays devil's advocate on risky trades, confirms before executing
Market Intelligence "how's the market today", "worth trading today" Broad market read — movers, funding rate landscape, overall risk-on/risk-off mood — independent of your own positions
Trading Journal (automatic, after every executed trade) Logs a structured, numbered entry — what executed, Guardian's verdict at the time, warnings raised — persisted in memory across sessions
Performance Analytics "what's my win rate", "why do I keep losing" Reads exchange trade history (source of truth for hard numbers) plus the journal (context for why) to surface real patterns, honestly — including saying "not enough data" when