Back to skill

Security audit

Chatmosp Parameter Builder

Security checks for vulnerabilities and agentic risk

Overview

This Markdown-only ChatMOSP parameter helper is purpose-aligned and requires user confirmation before calculations, with no evidence of hidden execution or data theft.

Before installing, note that the bilingual routing labels may cause wrong-language handling. Prefer the ClawHub companion skills, and verify any GitHub fallback source before using it.

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
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (8)

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 15)May include surrounding context.

md
> - If the user writes in Chinese, read `SKILL.md` in this same directory and respond in Chinese.

Hidden Instructions

High
Category
Prompt Injection
Confidence
60% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 1)May include surrounding context.

md
---
name: chatmosp-parameter-builder
description: |
  chatMOSP 系统的参数构建与管理中心。负责 MSR / KMC 任务的参数查询、智能补全、

Hidden Instructions

High
Category
Prompt Injection
Confidence
60% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · SKILL_cn.md (reported line 1)May include surrounding context.

md
---
name: chatmosp-parameter-builder
description: |
  chatMOSP 系统的参数构建与管理中心。负责 MSR / KMC 任务的参数查询、智能补全、

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The trigger definition is broad enough to activate this skill for many loosely related requests before calculation execution or whenever parameters are queried/modified. In an agentic system, ambiguous activation boundaries can cause unintended tool routing, parameter mutation, or interception of requests that should have been handled by a narrower skill, increasing the chance of unsafe workflow manipulation or incorrect downstream actions.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

These routing instructions are internally inconsistent and can deterministically send the agent to the wrong document based on user language. That is a prompt-integrity problem: an attacker could exploit language selection to steer execution toward a less restrictive or simply unintended instruction set, causing policy drift or incorrect downstream actions.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
98% confidence
Finding

The file-structure block says SKILL.md is the Chinese reference and SKILL_cn.md is 'This file (English)', but the analyzed file is SKILL.md and its content is clearly English. This directly conflicts with the inline documentation and the earlier routing instruction that says Chinese responses should read SKILL.md in the same directory.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
96% confidence
Finding

The language-routing block is self-contradictory: it says English users should read SKILL_cn.md, while Chinese users should continue using SKILL.md, but the current file content and later file-structure comments conflict about which file is Chinese versus English. In an agent system, contradictory authority rules can cause the wrong instruction file to be followed, producing inconsistent behavior and bypassing intended safeguards or workflow constraints.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
88% confidence
Finding

The file-structure section states SKILL.md is the Chinese file and SKILL_cn.md is the English file, which conflicts with the active document labeling and the earlier language-routing logic. This inconsistency increases the chance that operators or agents will load the wrong locale-specific instructions, undermining predictable behavior and weakening trust in the skill's control flow.

Content

No source excerpt is available for this finding.