Turning Your Trading Bot Into "Something the AI Can Call": MCP in a Quant System
MCP standardizes tool description, invocation and return shape. A practical guide to splitting meta / read / write tools, auth, description craft, tool-count limits and testing.
Related Content
Tags
The Real Culprit Behind CPU 100%: A Missing `ok` in Go and the Resulting Busy-Loop
Post-mortem of an intermittent CPU 100% — a Go select-read on a channel without ok, zero-values flood the loop after close. Includes SIGUSR1 dump trick and repo-wide sweep.
Picking an LLM Provider for Your Trading Bot: Gemini vs OpenAI vs Claude vs DashScope vs Kimi vs DeepSeek
Wiring an LLM into an automated trading system: route by task, abstract the provider, scrub secrets end-to-end. A field-tested guide across seven upstreams.
Related Posts
Picking an LLM Provider for Your Trading Bot: Gemini vs OpenAI vs Claude vs DashScope vs Kimi vs DeepSeek
Wiring an LLM into an automated trading system: route by task, abstract the provider, scrub secrets end-to-end. A field-tested guide across seven upstreams.
The Real Culprit Behind CPU 100%: A Missing `ok` in Go and the Resulting Busy-Loop
Post-mortem of an intermittent CPU 100% — a Go select-read on a channel without ok, zero-values flood the loop after close. Includes SIGUSR1 dump trick and repo-wide sweep.
Grid Trading Risk Control Dilemma and the Composite Risk Controller Solution
When multiple risk factors are simultaneously bearish but none reaches its individual trigger threshold, traditional independent risk checks fail. This article introduces QuantMesh's Composite Risk Controller — how it normalizes scattered signals, applies weighted aggregation for joint decision-making, and covers the ambiguous "cloudy day" risk scenarios in grid trading.