ClawHub
Back to skill

Security audit

sn-search-finance

Security checks across malware telemetry and agentic risk

Overview

This finance-search skill is coherent and disclosed, with expected network lookups and one documented user-directed download command.

Install only if you are comfortable with finance libraries making outbound network requests; review dependency versions if reproducibility matters, and use tdx-affair-fetch only with an intended download directory.

SkillSpector

Vulnerability Patterns
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (5)

Lp3

Medium
Category
MCP Least Privilege
Confidence
92% confidence
Finding
The skill explicitly instructs use of network-capable tools and external data sources such as yfinance, mootdx, API scripts, and browser-use, but the skill metadata shown here does not declare corresponding permissions. Undeclared network capability weakens security review and runtime consent because the skill can access external resources without a clear permission contract.

Description-Behavior Mismatch

Medium
Confidence
86% confidence
Finding
The skill presents itself primarily as a finance search helper, but it also exposes a command that downloads remote files to local storage. In an agent setting, undisclosed filesystem writes expand the trust boundary and can surprise users or higher-level orchestration that expected read-only data retrieval.

Context-Inappropriate Capability

Medium
Confidence
84% confidence
Finding
The code includes a local file download primitive via Affair.fetch using user-supplied filename and downdir, which goes beyond simple finance lookup. Even if intended for fetching finance datasets, this adds side effects and creates risk of unintended file creation or abuse when invoked by an autonomous agent.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
Unlike ordinary search commands, this command writes remote content into a local directory without any explicit warning in the interface. In an agent environment, silent filesystem modification is security-relevant because it can persist artifacts, consume storage, or place files in unexpected locations if inputs are not tightly controlled.

Unpinned Dependencies

Low
Category
Supply Chain
Content
yfinance
mootdx[all]
Confidence
94% confidence
Finding
The dependency 'yfinance' is specified without a version pin, which makes builds non-reproducible and allows future upstream releases to change behavior or introduce vulnerable or malicious code into the skill without review. In a finance-focused skill that may fetch market data or interact with external services, uncontrolled dependency updates increase supply-chain risk and can affect integrity, availability, or confidentiality depending on what the package version does.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.