Hotdog

Security checks across malware telemetry and agentic risk

Overview

The skill mostly matches its hot-dog photo game purpose, but it uploads photos to an external API using a hardcoded bearer token that is not safely scoped or declared.

Review this skill before installing. It is purpose-aligned for a hot-dog photo battle, but only use it with non-sensitive photos, and ask the publisher to remove the embedded bearer token or replace it with a properly scoped and declared credential mechanism.

VirusTotal

61/61 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.

#
ASI03: Identity and Privilege Abuse
High
What this means

Requests to the service are made under a shared embedded credential, which could be reused or abused and gives users no clear control over the API identity being used.

Why it was flagged

The skill contains a hardcoded bearer credential while the registry declares no required credentials or environment variables. The artifacts do not explain the token scope, ownership, or rotation.

Skill content
-H "Authorization: Bearer ih1rtmC7...XIi9X"
Recommendation

Do not ship a bearer token in the skill text; use a scoped server-side token or a declared user/configured credential with clear scope and rotation.

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

Food photos and the model's description/classification leave the local conversation and may be processed or stored by the external service.

Why it was flagged

The skill uploads the selected media file plus generated classification data to the stated hotdogornot API. This is purpose-aligned, but it moves user-provided image data to an external service.

Skill content
-F "image=@{{MediaPath}}" -F "claw_model={{Model}}" -F "claw_answer=ANSWER" -F "claw_reasoning=DESCRIPTION"
Recommendation

Use the skill only with photos you are comfortable sending to hotdogornot.xyz, and the publisher should clearly document retention and display behavior.

#
ASI02: Tool Misuse and Exploitation
Low
What this means

When invoked with a photo, the agent will make network requests and submit a vote automatically as part of the game flow.

Why it was flagged

The skill instructs the agent to run curl commands to post the round and submit a vote. This command execution is central to the stated battle workflow and is limited to the stated API endpoints.

Skill content
exec: curl -s -X POST "https://api.hotdogornot.xyz/api/arena/round"
Recommendation

Keep command execution limited to these endpoints, and prefer safely escaped arguments or a dedicated integration rather than manual shell-string substitution.