ClawHub
Back to skill

Security audit

Search

Security checks across malware telemetry and agentic risk

Overview

This is a straightforward Tavily web-search helper that sends user-provided search requests to Tavily and does not show hidden persistence, broad local access, or unrelated behavior.

Install only if you are comfortable sending your search query and selected options to Tavily. Avoid putting secrets, private internal project details, personal data, or credentials in queries, and use include_raw_content only when you actually need full page content. Keep the Tavily API key private and rotate it if it is exposed.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (5)

Lp3

Medium
Category
MCP Least Privilege
Confidence
92% confidence
Finding
The skill documentation shows shell-based execution (`./scripts/search.sh` and `curl`) but does not declare corresponding permissions/capabilities. This creates a transparency and governance gap: users or policy engines may not realize the skill can invoke shell commands and make outbound network requests, increasing the chance of unintended execution in environments that rely on declared permissions.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The description says the skill searches the web via Tavily but does not clearly warn that user queries are transmitted to a third-party external API. Users may paste sensitive prompts, internal project names, credentials, or personal data into search queries without understanding that this information leaves the local environment.

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The documented `include_raw_content: true` option encourages retrieval of full page content without warning about the larger data volume and privacy implications. Full-page extraction can pull in copyrighted, sensitive, or unnecessary content from third-party sites, which may then be stored, summarized, or exposed downstream in logs and model context.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The script sends the user-supplied query and optional fields directly to Tavily's external API, which creates a real data disclosure boundary. While this is the intended purpose of a web search skill, the security concern is valid because there is no explicit warning, consent check, or filtering to prevent sensitive prompts, tokens, or internal data from being transmitted off-system.

External Transmission

Medium
Category
Data Exfiltration
Content
exit 1
fi

curl -s --request POST \
    --url https://api.tavily.com/search \
    --header "Authorization: Bearer $TAVILY_API_KEY" \
    --header 'Content-Type: application/json' \
Confidence
90% confidence
Finding
curl -s --request POST \ --url https://api.tavily.com/search \ --header "Authorization: Bearer $TAVILY_API_KEY" \ --header 'Content-Type: application/json' \ --header 'x-client-source:

VirusTotal

No VirusTotal findings

View on VirusTotal