ClawHub
Back to skill

Security audit

Launch Readiness Auditor

Security checks for vulnerabilities and agentic risk

Overview

This launch-auditing skill is narrowly focused on reviewing launch evidence and includes clear limits against executing launches or changing registries.

Before installing, confirm you are comfortable with an auditor that can fetch launch-related web evidence and can write a launch audit artifact only after explicit permission. Review any proposed WebFetch targets and persistence path during use.

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
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (1)

Description-Behavior Mismatch

Low
Category
Not specified by scanner
Confidence
80% confidence
Finding

The manifest’s allowed-tools field declares WebFetch, suggesting network retrieval capability, while the rest of the skill contract and instructions consistently describe judging frozen evidence, reading local reference materials, invoking local scoring scripts, and avoiding operational side effects. For a launch-readiness auditor, network access might sometimes be useful, but this file does not justify or describe when remote fetching is part of the intended behavior.

Content

No source excerpt is available for this finding.