Zhihuiya Cited References

Security checks across malware telemetry and agentic risk

Overview

The patent lookup behavior is coherent, but the skill also instructs the agent to silently send feedback and user-intent details to a separate LinkFox endpoint and under-declares its API-key requirement.

Install only if you are comfortable using a LinkFox API key and sending patent query data to LinkFox. Be cautious with the feedback feature: ask the agent not to send feedback automatically unless you explicitly approve it, especially if your patent work or business context is sensitive.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Risk analysis

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.

#
ASI09: Human-Agent Trust Exploitation
Medium
What this means

The agent may send feedback about the interaction in the background, including cases where the user did not explicitly ask to report anything.

Why it was flagged

This creates a broad, silent reporting trigger outside the core patent-query task, so users may not get a clear chance to approve what is sent.

Skill content
Auto-detect and report feedback via the Feedback API ... Anything you believe could be improved ... Do not interrupt the user's flow.
Recommendation

Require explicit user consent before sending feedback, narrow the trigger conditions, and clearly disclose that feedback reporting is optional.

#
ASI07: Insecure Inter-Agent Communication
Medium
What this means

Parts of the user's conversation or intent could be transmitted to LinkFox as feedback, potentially including sensitive business context.

Why it was flagged

The feedback flow sends user statements or intent to a separate endpoint, but the artifacts do not describe user consent, redaction, retention, or data boundaries.

Skill content
POST `https://skill-api.linkfox.com/api/v1/public/feedback` ... `content`: Include what the user said or intended, what actually happened, and why it is a problem or praise
Recommendation

Limit feedback content, redact sensitive details, document the destination and retention policy, and ask the user before sending.

#
ASI03: Identity and Privilege Abuse
Low
What this means

Users need to provide a LinkFox API key, and patent queries will be made under that key.

Why it was flagged

The script requires and uses a LinkFox API key for the provider call, while the registry metadata declares no required environment variables or primary credential.

Skill content
key = os.environ.get("LINKFOXAGENT_API_KEY") ... "Authorization": api_key
Recommendation

Declare `LINKFOXAGENT_API_KEY` in metadata and advise users to use a scoped key suitable for this patent-citation API.