Skillv1.0.0

ClawScan security

jp-device-skill · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

SuspiciousMar 12, 2026, 8:35 AM

Verdict: suspicious
Confidence: medium
Model: gpt-5-mini
Summary: The skill claims vehicle-control and telemetry capabilities but is only a placeholder with no code, no required credentials, and no instructions for how to actually access devices — this mismatch is worth caution.
Guidance: This package is a placeholder: it advertises device control and telemetry for JinPeng vehicles but currently only replies 'skill under testing' and provides no code, API endpoints, or credentials. If you need a working integration, ask the publisher for the implementation details: which APIs or device gateways it will call, what credentials (device tokens, fleet API keys) are required, where network requests are sent, and whether any binaries or SDKs will be installed. Do not supply credentials until you can review the actual code or a trustworthy installation source; prefer skills that declare only the credentials they need (e.g., a single device API key) and that provide a clear install mechanism (official repo or release). If you install this as-is, it poses little direct risk because it does nothing, but it also provides no functionality — treat it as a stub and wait for a fully-specified release or request more information from the owner.
Findings: [no_code_no_findings] expected: The regex-based scanner found nothing because there are no code files. That is expected for an instruction-only placeholder skill, but it also means there is no implementation to review.

Review Dimensions

Purpose & Capability: concernThe SKILL.md describes connecting to JinPeng devices, reading telemetry, configuring parameters, and troubleshooting — actions that would normally require device APIs, network endpoints, and credentials. The package includes no code, no install steps, and declares no environment variables or credentials, so the described capabilities are not actually implemented here. The file also states the skill is 'under testing' and returns a fixed placeholder response, which explains the mismatch but leaves the capability unimplemented.
Instruction Scope: okThe runtime instructions are limited and safe: they specify when the skill should be used and then instruct the agent to reply with a fixed 'skill under testing' message. There are no commands, file reads, or network endpoints referenced in SKILL.md.
Install Mechanism: okThis is an instruction-only skill with no install spec and no code files. That minimizes installation risk because nothing is written to disk or downloaded.
Credentials: noteNo environment variables, credentials, or config paths are requested — which is consistent with the skill being a placeholder. However, the described device-control features would normally require credentials/APIs; the absence of any declared credentials is notable and means the skill as-published cannot perform its advertised functions.
Persistence & Privilege: okThe skill does not request always:true and uses platform defaults (user-invocable, model invocation allowed). It does not ask to modify other skills or system settings.