Skillv1.0.0

ClawScan security

可以登錄家電的保固，並且歸檔，提供查詢 · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

SuspiciousFeb 24, 2026, 10:22 AM

Verdict: suspicious
Confidence: medium
Model: gpt-5-mini
Summary: The packaged files and SKILL.md implement a 'Skill Creator / packager' tool, but the skill's name (in Chinese) and registry metadata advertise an unrelated appliance-warranty feature — this mismatch is unexplained and should be clarified before installing.
Guidance: Do not install or enable this skill until you confirm what it actually is. The code and SKILL.md clearly implement a 'skill authoring/packaging' utility (packager, validator, initializer), but the skill's public name/description advertise an unrelated appliance-warranty feature — this mismatch can be due to mislabeling, a packaging error, or malicious metadata. Actions to take before installing: - Verify the publisher/owner (kn75gf8ej0b4kptjgmpsqz6hgh81sd3q) and contact them or view the source repository to confirm intent. - Inspect SKILL.md and the scripts yourself (they are included) and confirm you are comfortable with a skill-authoring tool rather than a warranty service. - If you expected an appliance-warranty skill, refuse installation and ask the publisher for the correct package. If you intended a skill-authoring tool, consider renaming/rewriting frontmatter to match and confirm there are no hidden network behaviors. - Because the package contains executable scripts, run them in a sandboxed environment or review them line-by-line before execution. The code appears benign (no network or secret access), but the naming inconsistency is a red flag that warrants verification.

Review Dimensions

Purpose & Capability: concernRegistry name/title (Chinese: 'can register appliance warranties, archive, and provide query') claims an appliance warranty feature, but the SKILL.md frontmatter and body describe a 'Skill Creator' (tooling for building skills). The included scripts implement packaging/validation/init utilities for skills, which align with the SKILL.md but not with the user-visible skill name/description. This mismatch between advertised purpose and actual code is incoherent and could indicate mislabeling or accidental/intentional metadata tampering.
Instruction Scope: okSKILL.md instructions are confined to guidance for creating/structuring skills and best practices; they do not instruct reading arbitrary user files, secret environment variables, or sending data to external endpoints. The scripts (package, init, quick_validate) operate on local filesystem paths and frontmatter and do not perform network I/O or access credentials.
Install Mechanism: okNo install spec is provided (instruction-only skill with bundled scripts). No remote downloads, package installs, or archive extraction from external URLs are present in the manifest or scripts — this is low-risk from an install mechanism perspective.
Credentials: okThe skill declares no required environment variables, credentials, or config paths. The code likewise does not reference environment secrets. Requested privileges/credentials are proportionate (none) to the actual local-filepackaging purpose of the scripts.
Persistence & Privilege: okalways is false and the skill does not request persistent system-wide presence or attempt to modify other skills' configs. The agent-invocation defaults are normal. Bundled scripts create files when initializing a template but do not alter system/global settings.