v1.1.8

PIPL-Compliance（PIPL合规工具）

BenignClawScan verdict for this skill. Analyzed May 1, 2026, 8:21 AM.

Analysis

This appears to be a local PIPL compliance toolkit with user-directed Python commands; no artifact-backed exfiltration, hidden network use, or destructive behavior was found.

GuidanceThis skill looks purpose-aligned for local PIPL compliance work. Before installing, use a virtual environment, review dependency versions, run it first on sample or anonymized data, protect generated reports, and do not rely on it as legal advice or as a substitute for professional review.

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.

Abnormal behavior control

Checks for instructions or behavior that redirect the agent, misuse tools, execute unexpected code, cascade across systems, exploit user trust, or continue outside the intended task.

Agentic Supply Chain Vulnerabilities

SeverityLowConfidenceHighStatusNote

requirements.txt

pandas>=2.0.0
jinja2>=3.1.0

The skill relies on external Python packages specified with version lower bounds rather than exact pinned versions. This is normal for a Python CLI tool, but dependency resolution can change over time.

User impactInstalling the dependencies may pull newer package versions than the author tested.

RecommendationInstall in a virtual environment, review the resolved package versions, and pin versions in your own environment if reproducibility matters.

Tool Misuse and Exploitation

SeverityLowConfidenceHighStatusNote

SKILL.md

pip install -r requirements.txt
python scripts/pipl-check.py --scenario user-registration --output report.json

The skill asks users or agents to run local Python commands and write report files. This is central to its stated purpose and is user-directed, but it is still local command/file-write capability.

User impactThe tool can create or overwrite local report/document files in the working directory or chosen output paths.

RecommendationRun commands from a dedicated project directory, avoid overwriting important files, and review generated output before using it for business decisions.

Human-Agent Trust Exploitation

SeverityInfoConfidenceHighStatusNote

SECURITY_CHECK_GUIDE.md

总体评估 ... 可以安全安装使用

The guide includes self-attested security conclusions. They are accompanied by suggested checks and are not contradicted by the provided static scan, but users should not treat self-attestation as independent approval.

User impactA user might over-rely on the package's own safety claims instead of reviewing the code, dependencies, and generated outputs.

RecommendationTreat the included safety guide as helpful documentation, not a guarantee; verify dependency sources and consult security/legal professionals for high-stakes use.

Sensitive data protection

Checks for exposed credentials, poisoned memory or context, unclear communication boundaries, or sensitive data that could leave the user's control.

Memory and Context Poisoning

SeverityLowConfidenceHighStatusNote

scripts/utils/data_validator.py

'id_card': '身份证号码',
'phone': '手机号码',
'email': '电子邮箱',
'address': '住址',
'biometric': '生物识别信息'

The validator is designed to inspect personal and sensitive information fields, and the workflow generates local reports/documents that may preserve compliance inputs or summaries.

User impactReal personal information or sensitive compliance details could remain in local generated files or conversation context if the user supplies them.

RecommendationUse sample or anonymized data when possible, store generated reports securely, and avoid pasting unnecessary real personal information into agent prompts.