ClawHub
Back to skill

Security audit

Recraft AI

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed Recraft API image tool with a user-invoked account-info command that deserves privacy awareness but does not show hidden or malicious behavior.

Install only if you are comfortable giving the skill a Recraft API token and sending selected prompts or images to Recraft. Avoid confidential images unless Recraft is approved for that data, choose output paths carefully, and avoid running user-info where account email or credit details may be captured in shared logs.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Tp4

High
Category
MCP Tool Poisoning
Confidence
83% confidence
Finding
The skill description focuses on image generation and editing, but the documented `user-info` command expands behavior into account-data retrieval, including user ID, name, email, and credits. That mismatch can mislead users and reviewers about the skill's actual data-handling scope, increasing the risk of unintended exposure of personal or account metadata.

Description-Behavior Mismatch

Medium
Confidence
95% confidence
Finding
The script advertises image-processing functions but its usage text also exposes a `user-info` command, expanding capability into account-data retrieval outside the stated purpose. This mismatch can surprise users and enables collection of profile data with the same API token, which is a scope-creep/privacy risk even if no exploit primitive is present.

Description-Behavior Mismatch

Medium
Confidence
98% confidence
Finding
The `cmd_user_info` function calls `/users/me` and prints account attributes unrelated to image generation or editing. Pulling profile information, especially when not disclosed by the skill metadata, increases privacy risk and broadens token use beyond the user's likely expectation.

Context-Inappropriate Capability

Medium
Confidence
99% confidence
Finding
The script displays account metadata including email, name, ID, and credits directly to stdout. In shared terminals, logs, or agent environments, this can unnecessarily expose personal or account-identifying information unrelated to the declared image-processing function.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.