Back to skill
Skillv1.0.0
ClawScan security
Baostock 1.0.3 · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignMar 29, 2026, 4:29 AM
- Verdict
- benign
- Confidence
- high
- Model
- gpt-5-mini
- Summary
- This skill is internally consistent with its stated purpose (a Python-based BaoStock data helper); it only requires python3 and uses the public baostock/pandas libraries and contains no unexpected credential or network-exfiltration requests.
- Guidance
- This skill appears coherent and limited to fetching A‑share data via the public baostock Python library. Before installing or running: 1) confirm you trust the baostock PyPI package and its homepage (https://www.baostock.com) and review the upstream package source if you need higher assurance; 2) install and run in a virtualenv or isolated environment; 3) be aware the code will make network requests to BaoStock endpoints when you call bs.login()/query_*, so only run it where network access is acceptable; 4) note the minor version/metadata inconsistencies in the bundle — they look like bookkeeping issues but you can verify the package version on PyPI if version strictness matters.
Review Dimensions
- Purpose & Capability
- okName/description (BaoStock A股 data) match the included instructions, demo, and required binaries. The skill only depends on python3 and references the public baostock and pandas packages, which are appropriate for the stated functionality. (Minor metadata/version inconsistencies across files (_meta.json, SKILL.md, metadata.json) are present but do not indicate malicious behavior.)
- Instruction Scope
- okSKILL.md and the demo instruct the agent to pip-install baostock, call bs.login()/bs.logout(), and run query_* functions; there are no instructions to read unrelated files, access credentials, or send data to unexpected external endpoints. All runtime actions are within the declared purpose (fetching market data).
- Install Mechanism
- okThis is instruction-only (no install spec). Dependencies are standard PyPI packages (baostock, pandas) referenced in requirements.txt; there are no downloads from arbitrary URLs or extracted archives in the skill bundle.
- Credentials
- okThe skill requests no environment variables or credentials. Demo uses bs.login() without requiring API keys, consistent with BaoStock's public/no-key model. No unexpected secret access is requested.
- Persistence & Privilege
- okSkill does not request always:true, does not modify other skills or system-wide settings, and allows normal autonomous invocation. No elevated persistence or cross-skill configuration is present.
