Skillv1.0.1

ClawScan security

新闻事实核查 · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

BenignMar 11, 2026, 8:36 AM

Verdict: benign
Confidence: high
Model: gpt-5-mini
Summary: This instruction-only skill is internally consistent with a news fact‑checking assistant: it asks for no secrets, installs nothing, and its runtime instructions stay on topic.
Guidance: This skill appears coherent and low‑risk: it is a set of procedures and resource links for fact checking and asks for no secrets or installs. Before installing, consider: (1) the skill has no published homepage/source — if provenance matters, prefer skills with a clear publisher or reviews; (2) test it with non‑sensitive example news to confirm the outputs meet your standards; (3) the agent/platform may still need internet access to follow the referenced sites — avoid submitting private or sensitive content for online verification; (4) the AIGC metadata in the SKILL.md looks like provenance labels only, but if you need guarantees about origin, ask the publisher for more information.

Purpose & Capability: okName/description (news fact checking) match the SKILL.md and references: all guidance and tool suggestions are appropriate for verifying news items. No unrelated credentials, binaries, or config paths are requested.
Instruction Scope: noteSKILL.md contains only procedures, checklists, and links to public fact‑checking sites and image/video verification tools — all on‑topic. There is a small metadata block (AIGC/ProduceID/ReservedCode fields) in the file header; it appears to be provenance/labeling metadata and does not instruct any off‑scope actions, but its presence is informational rather than functional.
Install Mechanism: okNo install spec and no code files — nothing will be written to disk or fetched at install time. Instruction-only skills are lowest install risk.
Credentials: okThe skill requires no environment variables, credentials, or config paths. All referenced external resources are public fact‑checking sites and verification tools appropriate for the stated purpose.
Persistence & Privilege: okalways is false and there is no request for persistent system changes. disable-model-invocation is false (agent may invoke it autonomously), which is the platform default and not by itself a concern given the skill's limited scope.