Skillv1.0.0

ClawScan security

PyWenCai Stock · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

BenignMar 7, 2026, 12:02 PM

Verdict: benign
Confidence: high
Model: gpt-5-mini
Summary: The skill's code, instructions, and requirements are consistent with its stated purpose of providing pywencai-based A‑share data queries; no disproportionate credentials, installs, or unexpected behaviors were found.
Guidance: This skill appears to do what it says: wrap the third‑party pywencai library to fetch A‑share data. Before installing, consider the following: - Review and trust the upstream pywencai package (pip). It will make network requests to retrieve data; inspect its code or pin a specific version if you need stability/trust. - Do NOT apply the SSL-disable snippet from the troubleshooting section in production; it makes HTTPS connections insecure and can enable man‑in‑the‑middle attacks. - The example script adds '/tmp/mootdx' to sys.path — harmless by itself, but avoid running arbitrary example scripts that modify import paths in sensitive environments. - If you must supply login cookies or any credentials to access additional data, treat them as sensitive: provide them only in secure ways, prefer temporary/session tokens, and avoid embedding them in code. - Run initial tests in an isolated environment (virtualenv/container) to observe network behavior and confirm no unexpected endpoints are contacted. Overall the package is internally consistent with its stated purpose; the main residual risk is trusting the third‑party pywencai package and following the insecure SSL workaround.

Review Dimensions

Purpose & Capability: okThe name/description (pywencai access to 同花顺问财) matches the included code and docs. The code imports pywencai/pandas/BeautifulSoup and exposes search/top_gainers/dragon_tiger_list which aligns with advertised features. Minor oddity: scripts/example.py inserts '/tmp/mootdx' on sys.path — likely a convenience for some environments but not required for core functionality.
Instruction Scope: noteSKILL.md stays within scope: it instructs installing pywencai/pandas and shows how to call skill('pywencai-stock').search. It does include a troubleshooting snippet that disables SSL verification (ssl._create_default_https_context = ssl._create_unverified_context), which is insecure advice — functionally coherent but a security risk if followed. Instructions do not request extraneous files, system credentials, or unexpected external endpoints.
Install Mechanism: okNo install spec is provided (instruction-only). Recommended installs are standard pip packages (pywencai, pandas). No downloads from untrusted URLs, no extracted archives, and no unusual install locations in the core package files.
Credentials: okThe skill declares no required environment variables or credentials. SKILL.md notes some queries may need login cookies (a legitimate property of the upstream service) but does not demand unrelated secrets. There is no request for broad or unrelated credentials.
Persistence & Privilege: okSkill does not request always:true and is user-invocable only. It contains no code that modifies other skills or global agent configuration. Autonomous invocation remains enabled (platform default) but is not combined with other red flags.