ClawHub
Back to skill

Security audit

open-claude

Security checks for vulnerabilities and agentic risk

Overview

This skill is a simple browser launcher for Claude in Brave, with some rough implementation details but no evidence of hidden data access, persistence, exfiltration, or destructive behavior.

Before installing, understand that invoking this skill may open Brave and connect to Claude.ai. The publisher should narrow the trigger phrases, add an explicit tool/permission declaration, and replace the Windows shell=True launch with a safer browser-opening method.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (7)

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
94% confidence
Finding

This duplicate finding points to the same underlying issue: shell-mediated launching of an external tool on Windows. In an agent skill context, launching external applications is more sensitive because future reuse with user-supplied URLs or browser names could enable abuse beyond the current hardcoded case.

Content

Scanner excerpt · scripts/open_claude.py (reported line 9)May include surrounding context.

python
system = platform.system()

if system == "Windows":
    subprocess.Popen(["start", "brave", url], shell=True)
elif system == "Darwin":
    subprocess.Popen(["open", "-a", "Brave Browser", url])
else:

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
94% confidence
Finding

This duplicate finding points to the same underlying issue: shell-mediated launching of an external tool on Windows. In an agent skill context, launching external applications is more sensitive because future reuse with user-supplied URLs or browser names could enable abuse beyond the current hardcoded case.

Content

Scanner excerpt · scripts/open_claude.py (reported line 9)May include surrounding context.

python
system = platform.system()

if system == "Windows":
    subprocess.Popen(["start", "brave", url], shell=True)
elif system == "Darwin":
    subprocess.Popen(["open", "-a", "Brave Browser", url])
else:

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

Several trigger phrases are broad enough to activate unintentionally, especially phrases like "start claude," "claude please," or "open anthropic." Accidental invocation could launch a browser unexpectedly and navigate the user to an external service, which is not severe but is still undesirable and could erode trust or create minor workflow and privacy issues.

Content

No source excerpt is available for this finding.

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
95% confidence
Finding

The Windows branch invokes subprocess.Popen with shell=True, which routes execution through the shell and increases risk if any argument ever becomes attacker-controlled. In this file the URL is hardcoded, so there is no immediate injection path, but the pattern is still unsafe and unnecessary for opening a browser.

Content

Scanner excerpt · scripts/open_claude.py (reported line 9)May include surrounding context.

python
system = platform.system()

if system == "Windows":
    subprocess.Popen(["start", "brave", url], shell=True)
elif system == "Darwin":
    subprocess.Popen(["open", "-a", "Brave Browser", url])
else:

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/open_claude.py (reported line 11)May include surrounding context.

python
if system == "Windows":
    subprocess.Popen(["start", "brave", url], shell=True)
elif system == "Darwin":
    subprocess.Popen(["open", "-a", "Brave Browser", url])
else:
    subprocess.Popen(["brave-browser", url])

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/open_claude.py (reported line 13)May include surrounding context.

python
elif system == "Darwin":
    subprocess.Popen(["open", "-a", "Brave Browser", url])
else:
    subprocess.Popen(["brave-browser", url])

print("✅ Claude opening in Brave!")

Static analysis

No suspicious patterns detected.