@lanyasheng

trading-quant provides real-time stock quotes and multi-dimensional scoring across cn, us, hk markets and commodities via multiple data sources. lacks actionable procedures and error handling specifics.

auto-improvement-orchestrator-skill provides a pipeline for evaluating skills across 9 dimensions and auto-generating improvements via multi-stage gate filtering. lacks concrete procedural steps and failure mode documentation.

trading-quant wraps chinese market data sources (tencent, sina, eastmoney, tonghuashun) for real-time quotes and multi-dimensional scoring. skill definition is incomplete and lacks actionable procedural detail.

AI Agent 的网页访问路由决策指南。全 CLI 架构，零 MCP 依赖，不占常驻上下文 token。 按成本逐级升级: WebFetch($0) → opencli web read($0,带Cookie) → Firecrawl → agent-browser → browser-use。 IMPORTA...

当需要对改进候选多人盲审打分、用 LLM 做语义评估、判断候选是否应被接受、或打分结果全是 hold 想知道为什么时使用。支持 --panel 多审阅者盲审和 --llm-judge 语义评估。不用于结构评估（用 improvement-learner）或门禁决策（用 improvement-gate）。

Agent 错误恢复与容错。当 session 遇到限速、crash 或模型失败时使用。不适用于工具重试死循环（用 tool-governance）或 agent 提前停止（用 execution-loop）。参见 tool-governance（错误追踪数据）。