ClawHub
Back to skill

Security audit

Perplexity Search Skill

Security checks across malware telemetry and agentic risk

Overview

This is a straightforward Perplexity web-search skill that discloses its API key use and external request behavior, with the main caveat that searches are sent to a third-party service.

Install this only if you are comfortable sending search queries to Perplexity and using your Perplexity API quota or billing. Avoid including secrets, private documents, customer data, internal URLs, or confidential project details in searches unless sharing that information with Perplexity is acceptable.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • 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)

Lp3

Medium
Category
MCP Least Privilege
Confidence
88% confidence
Finding
The skill documentation indicates the capability to access an environment variable and make outbound network requests, but there is no explicit permission declaration governing those actions. In agent ecosystems that rely on declared permissions for policy enforcement and user consent, this mismatch can lead to under-scoped review and unintended execution of sensitive capabilities.

Vague Triggers

Medium
Confidence
89% confidence
Finding
The README states the skill is auto-invoked on "web-search intent," which is a broad and underspecified trigger. In an agent environment, vague activation criteria can cause unintended transmission of user prompts or sensitive context to an external service when the user did not clearly consent to web access.

Vague Triggers

Medium
Confidence
93% confidence
Finding
The usage section says the agent invokes this skill automatically when asked about "anything web-related," which is excessively broad. This increases the chance that normal conversation, sensitive research, or mixed-intent prompts are sent to Perplexity without sufficiently specific user authorization.

Vague Triggers

Medium
Confidence
91% confidence
Finding
The trigger phrases are broad enough to match many ordinary user requests such as "current," "recent," or "find out," which increases the chance of automatic invocation without clear user intent. Because this skill performs network access and uses a secret from the environment, overbroad activation can cause unnecessary external data transmission and surprise tool use.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.