ClawHub
Back to skill

Security audit

Virtual Reading Group

Security checks for vulnerabilities and agentic risk

Overview

This markdown-only skill coherently sets up a user-directed academic reading group workflow, with expected file outputs and no evidence of hidden execution, credential access, persistence, or exfiltration.

Installers should expect this skill to read the papers they provide, run several agent passes, and create multiple markdown files. Use a new dedicated output directory for each run, avoid placing sensitive papers in shared folders, and confirm costs before using high-end models or large paper sets.

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
  • 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 (3)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The manifest description uses very broad trigger phrases like analyzing literature, running paper discussions, and synthesizing research across multiple sources, which can cause the skill to activate for many ordinary academic-assistance requests. Over-broad invocation increases the chance the agent selects this skill unnecessarily, leading to unexpected multi-agent orchestration, extra cost, and unintended file creation or processing of user documents.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill explicitly writes many output files and may create the output directory, but the high-level description does not clearly disclose these filesystem side effects before execution. Users may invoke it expecting analysis only, while the skill performs persistent writes across multiple phases, which can surprise users, overwrite expected workspace contents, or expose sensitive paper notes in shared directories.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
90% confidence
Finding

The workflow repeatedly instructs agents to save outputs into {output_dir} but does not require confirmation, uniqueness checks, or safeguards against overwriting existing files. In an orchestration context, a reused or attacker-influenced output directory could cause clobbering of prior data, loss of user files, or unintended modification of artifacts consumed by later phases.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.