Quantlib Derivatives

通过 SWIG 绑定调用 QuantLib 引擎，完成期权、互换、债券等金融衍生品的定价计算，支持美式期权有限差分法和篮子价差期权等多资产策略验证。。

This skill appears to be a knowledge-rich, instruction-only wrapper for QuantLib/SWIG and ZVT-based backtests, but there are important mismatches you should consider before installing or running it: - Environment: The SKILL.md requires Python 3.12+, SWIG/QuantLib SWIG bindings and ZVT, but the skill declares no install steps. Ensure your environment already has QuantLib, the SWIG Python bindings, and zvt installed (or be prepared to install them manually) before use. - Credentials: The skill references data providers (joinquant, eastmoney, akshare, qmt). Some providers (e.g., joinquant) require API keys/accounts. Expect the agent to request or need credentials at runtime — the skill did not declare these, so plan to provide them only after verifying the author/source. - Filesystem actions: The instructions include running precondition checks that read and write to a ZVT home directory (default ~/.zvt). If you run this skill, do so in a sandbox or ensure it has access only to directories you are comfortable with. - No code install vs host expectations: Because there is no install recipe, the skill will either fail when dependencies are missing or the agent may attempt to run pip/recorder commands during preconditions. Confirm who will run those commands and whether network access is allowed. - Verification steps: Ask the publisher for an explicit dependency/install list and for a manifest of any runtime network endpoints the skill will call. If you cannot verify the author, run initial experiments in an isolated environment (container/VM) and monitor file and network activity. What would change this assessment: presence of an explicit, trusted install spec (with known package sources), declared required env vars/credential names, or included code that transparently shows only local computations and calls to known provider APIs would increase confidence. Conversely, any hidden network endpoints, undeclared credential usage, or instructions to run arbitrary shell/python code would increase concern.

Type: OpenClaw Skill Name: quantlib-derivatives Version: 0.3.3 The skill bundle is a highly structured framework for quantitative finance and derivatives pricing using QuantLib and the ZVT library. It contains extensive domain-specific constraints, semantic locks (e.g., SL-01 to SL-12), and validation rules designed to prevent common quantitative errors like look-ahead bias and incorrect financial conventions. The instructions in SKILL.md and seed.yaml are focused on enforcing these financial rules and managing the agent's persona. No evidence of data exfiltration, malicious execution, or harmful prompt injection was found; the installation recipes and preconditions are limited to standard environment checks and the installation of the legitimate 'zvt' package.