Yfinance Market Data

通过 Yahoo Finance 获取全球多市场股票、指数、外汇及加密货币的历史行情、财务数据、实时报价和财务日历。

This skill is broadly coherent with a data-pipeline/backtest helper, but has several practical mismatches you should be aware of before installing or running it: - The skill will check for and may ask you to install Python packages (zvt) and initialize a data directory (~/.zvt). Only proceed if you trust the upstream packages and are comfortable the agent may run pip or shell commands. - SKILL.md reads the ZVT_HOME environment variable and expects provider API tokens (joinquant, etc.) though none are declared in the registry metadata. Do not provide sensitive API keys without verifying the skill's source and license. Prefer creating minimal-permission keys or sandbox accounts for testing. - The skill references an unfamiliar 'uv' package manager; verify what that is (and its trustworthiness) before installing or allowing any automatic install steps. - Evidence-quality notes in SKILL.md warn that only ~30% of evidence was verified and show audit failures — treat generated results as potentially incomplete and validate critical outputs against authoritative sources. - Practical steps: ask the publisher/source (there is none listed) for a homepage or upstream repo; verify the LICENSE and the zvt package versions it intends to use; set ZVT_HOME to a writable directory you control; run any install steps in a disposable environment (virtualenv/containers) first; and avoid pasting high-privilege credentials into prompts until you confirm the skill's provenance. If the publisher/source is provided and the skill's metadata is updated to declare the environment requirements (ZVT_HOME, provider tokens) and explain the 'uv' manager, my confidence that this is benign would increase.

Type: OpenClaw Skill Name: yfinance-market-data Version: 0.3.3 This skill bundle provides a highly structured framework for financial market data retrieval and backtesting, primarily utilizing the `zvt` library and Yahoo Finance. The instructions for the AI agent (found in `SKILL.md` and `seed.yaml`) are focused on enforcing financial logic and data integrity, such as 'Semantic Locks' to prevent look-ahead bias (`SL-02`) and 'Constraints' to ensure proper HTTP session management (`finance-C-001`). There is no evidence of data exfiltration, malicious code execution, or harmful prompt injection; the bundle's complexity is dedicated to preventing common quantitative trading errors and ensuring compliance with market-specific rules like A-share T+1 trading.