量化策略研发实验室
PassAudited by ClawScan on May 5, 2026.
Overview
The skill is a disclosed, instruction-only quant research workflow, but users should notice its persistent local state, multi-agent context sharing, external install instructions, and high-stakes financial framing.
This appears safe to install as an instruction-only research skill if you understand that it stores pipeline state locally and may pass your market, capital, focus, and project context into subagent prompts. Do not provide broker credentials or secrets, verify any external repository before installing from GitHub, and treat generated trading strategies as educational research rather than financial advice.
Findings (4)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
Past strategy outputs and financial context can influence later runs, and sensitive user-entered details may remain in the project directory.
The skill creates and reuses a persistent local research-context file containing user-provided market, capital, and focus information.
检查 `state/research-context.md` 是否存在... 新建... 将答案填入 state 文件的 `## 项目上下文` 区
Review or reset the state file between projects, and do not enter exchange credentials, account numbers, or other secrets into the workflow.
User-provided financial context and project architecture notes may be included in subagent prompts during pipeline execution.
The skill passes collected context and role prompts into subagents as part of its multi-agent workflow.
Dispatch subagent：使用 Agent tool，`subagent_type: "general-purpose"`... 注入替换后的完整 prompt
Use the pipeline only for information you are comfortable sharing with the agent runtime, and avoid including private credentials or confidential account details.
If installing from the external repository rather than the reviewed package, the installed content could differ from the provided artifacts.
The README documents installation from an external GitHub repository via npx; this is user-directed and not an automatic install step in the skill package.
npx skills add https://github.com/yili1992/quant-research-lab --skill quant-research-lab
Install from a trusted source, review the repository contents, and pin a known commit or version when possible.
Users could overvalue generated trading strategies because of the institutional-style framing, even though the artifact says it is educational only.
The skill uses strong institutional branding for financial strategy generation while also clearly disclaiming that it is not trading advice.
institutional-grade quantitative strategy research workflow... modeled after the expertise of a top quant firm... Disclaimer: This framework is designed for research and educational purposes only.
Treat outputs as research drafts, validate independently with real data and professional review, and do not trade solely based on generated results.