ToolProduction ReadyMCPOPENAI-FUNCTIONS

PanWatch

Name: PanWatch
Rating: 3.2 (2212 reviews)
Author: TNT-Likely

by TNT-Likely

AI-driven stock monitoring and portfolio alerts across A/HK/US markets

Python

Updated Oct 23, 2026

2.2k

Stars

401

Forks

Commits/Week

Commits/Month

View on GitHub

What It Does

Monitors A-share, HK and US stock markets with AI agents that generate portfolio insights, alerts, and automated reports. Combines data sources (akshare, market APIs) with agent workflows implemented via langgraph/MCP and LLMs to run continuous monitoring and signal generation. Exposes real-time alerts and report endpoints (FastAPI) so trading workflows and UIs can subscribe to agent outputs. Orchestrator-Worker Pattern and Supervisor Pattern help coordinate and supervise agent workflows across markets.

Explore evaluation patterns

Methods to assess this tool

Key Benefits

As trading systems automate decisions, knowing which agent signals are reliable matters for money and risk management. PanWatch centralizes agent-generated signals and traces them to data and agent steps, making it easier to spot recurring failure modes or overconfident models. That visibility helps teams build track records and governance around agent-driven trading signals. Reflection Pattern and Emergence-Aware Monitoring Pattern provide perspectives on how signals evolve and how to monitor for emergent behavior.

When to Use

Quant teams and fintech builders who need continuous agent monitoring, automated reports, and signal traceability for production trading workflows. Leverage Event-Driven Agent Pattern for reactive monitoring and Planning Pattern to coordinate multi-step trading actions.