Back to skill
v1.0.0
data-scientist
BenignClawScan verdict for this skill. Analyzed Apr 30, 2026, 12:43 PM.
Analysis
This is a low-risk, instruction-only data science helper with no declared installs, credentials, code files, or autonomous system access.
GuidanceThis appears safe to install as an instruction-only data science skill. As with any generated analysis or notebook code, review it before running it, confirm any optional library or cloud-service use yourself, and avoid sharing sensitive datasets unless necessary.
Findings (1)
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
SeverityInfoConfidenceHighStatusNote
metadata
Source: unknown; Homepage: none
The skill has limited provenance information. This is not a material concern here because the artifact is instruction-only with no install spec or code files, but users have less information about maintainership.
User impactYou have limited external context for who maintains the skill, although there is no automatic installation or executable component in the provided artifacts.
RecommendationReview the instructions and examples before relying on them, and prefer skills with clearer provenance for high-trust or regulated work.