ClawHub

Ptengine Heatmap Analyze

Security checks across malware telemetry and agentic risk

Overview

The skill’s heatmap-analysis purpose is coherent, but it needs review because it tells the agent to handle a Ptengine API key despite saying it does not.

Review before installing. Do not paste a Ptengine API key into the agent chat; configure ptengine-cli manually or through a secure credential prompt, use a least-privileged key if possible, and remember that heatmap queries may expose sensitive business analytics.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Intent-Code Divergence

High
Confidence
98% confidence
Finding
The skill claims it never reads, writes, or transmits the API key itself, yet later instructs the agent to ask the user for the API key and invoke `ptengine-cli config set --api-key <KEY> ...`. That means the agent necessarily handles secret material in prompt/tool context, creating risk of credential exposure through logs, transcripts, command history, or downstream tool telemetry.

Intent-Code Divergence

Medium
Confidence
95% confidence
Finding
The disclosure states `config.yaml` is read only by `ptengine-cli`, but the workflow explicitly tells the skill to check whether the file exists and contains `api_key:`. This is a direct contradiction that expands the skill's access to credential-adjacent filesystem data and weakens user trust about what the skill inspects.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal