ClawHub
Back to skill

Security audit

Comm LITE

Security checks for vulnerabilities and agentic risk

Overview

This is a read-only Markdown helper for drafting communication replies, but its documentation includes confusing boilerplate about capabilities it does not actually have.

Install this only as a drafting and reply-polishing guide. Do not treat its boilerplate references to file writing, APIs, networking, or message delivery as permission for the agent to modify files, call external services, handle credentials, or send messages without separate user-directed approval.

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 (6)

Intent-Code Divergence

High
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill is presented as a read-only Markdown communication helper, but later claims capabilities such as file writing and external API integration that exceed the declared tool scope. This mismatch can mislead an agent or operator about what actions are appropriate, increasing the risk of unsafe tool routing, overbroad trust, or accidental privilege expansion if the skill is later paired with more capable tools.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The description says to use the skill whenever there is message sending, notification pushing, email/SMS, or communication integration, which is a very wide activation scope without concrete trigger phrases or exclusion boundaries. This ambiguity increases the chance of unintended invocation for ordinary communication-related requests.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The file states '支持中文交互' and presents the skill as a Chinese-interaction tool, but does not indicate that language selection is optional or user-controlled. This can violate language/locale policy expectations when a skill defaults to a specific language without explicit opt-in.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The core feature summary explicitly describes the skill as providing Chinese interaction, with no adjacent language preference mechanism or opt-in wording. In the absence of a documented user choice, this appears to impose a language default.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The error-handling section instructs users to run ping and troubleshoot firewall/proxy settings, which is inconsistent with a read-only Markdown skill that declares only the read tool. In an agent ecosystem, this kind of operational overreach can cause the model to suggest or attempt commands outside its intended scope, undermining least-privilege expectations and increasing the chance of unsafe command execution in broader environments.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The documentation states that no extra API keys or external communication services are required, yet the security and networking sections discuss API-key handling, HTTPS usage, and SSL validation. These contradictions create an unreliable security model and may normalize unnecessary secret handling or network assumptions in a skill that should be local and instruction-only.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.