ClawHub
Back to skill

Security audit

Stream

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed decision-making assistant with only limited, optional project-file saving and no evidence of hidden or unsafe behavior.

Installers should expect an interactive decision wizard that may activate on broad decision-making phrases. Only agree to save a journal if you are comfortable storing the decision summary in the current project's docs/decisions/ directory.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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
Findings (2)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The trigger phrases include generic intents like 'help me decide' and 'pros and cons', which are broad enough to cause accidental invocation in unrelated contexts. Unintended activation matters here because the skill can ask multiple guided questions and optionally write files, increasing the chance of workflow disruption or user confusion if selected when another tool or skill was more appropriate.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Low
Category
Not specified by scanner
Confidence
90% confidence
Finding

The manifest description frames the skill as a conversational STREAM analysis tool for helping founders make decisions. In the skill instructions, it also offers to save a decision journal by writing into docs/decisions/, which is an additional persistence behavior not reflected in the high-level description.

Content

No source excerpt is available for this finding.