Skillv0.1.1

ClawScan security

3dgs Method Compare · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

BenignApr 30, 2026, 10:46 AM

Verdict: benign
Confidence: high
Model: gpt-5-mini
Summary: This is an instruction-only skill that provides comparative analyses of 3D Gaussian Splatting methods; it requests no credentials, makes no system calls or installs, and its declared behavior matches what it asks for.
Guidance: This skill appears coherent and low-risk: it's an instruction-only knowledge/analysis engine for comparing 3D Gaussian Splatting methods and asks for nothing sensitive. Before relying on any specific claims (paper venues, metrics, compression ratios, or '105+ methods'), verify citations and original papers — instruction-only skills can confidently present plausible-sounding but incorrect or dated details. Avoid pasting secrets or private files into prompts; if you need reproducible code or implementations, ask the skill to cite links to official repos/papers and then inspect those sources yourself.

Purpose & Capability: okThe name and description (method comparison for 3D Gaussian Splatting) align with the SKILL.md content: it defines comparison dimensions, known-method tables, and guidance for producing publication-quality comparisons. There are no unrelated resource requests (no env vars, no binaries, no config paths) that would be inconsistent with a research/comparison skill.
Instruction Scope: okSKILL.md instructs the agent to act as an expert and produce multi-dimensional comparisons and tables; it does not instruct reading files, accessing system configuration, exfiltrating data, or calling external endpoints. Triggers are natural-language phrases for activating the skill. Nothing in the provided instructions grants the skill broader system access or asks for secrets.
Install Mechanism: okNo install specification and no code files — the skill is instruction-only. This is the lowest-risk model because nothing is written to disk or executed beyond the agent following prose instructions.
Credentials: okThe skill declares no required environment variables, credentials, or config paths. There is no disproportionate credential request relative to the stated purpose.
Persistence & Privilege: okalways is false and the skill does not request elevated persistence or system-wide changes. It is user-invocable and can be autonomously invoked by the agent per platform defaults, which is normal for skills of this type.