ClawHub
Back to skill

Security audit

命题工坊

Security checks for vulnerabilities and agentic risk

Overview

This is a markdown-only skill for generating algorithm contest programming problems, with no hidden execution or sensitive access behavior found.

Installers should expect this skill to steer relevant requests toward OI/CSP-style contest problem generation. If you only want it for formal contest authoring, narrow the generic trigger wording, but the inspected artifact does not show hidden code execution or sensitive data access.

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 (2)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The trigger list includes very broad terms such as “编程题”, which can match many generic requests for programming problems rather than the narrowly intended OI/NOI/CSP contest-setting use case. Over-broad activation can cause the skill to take over unrelated requests and apply highly specific constraints or behaviors outside its intended scope, increasing the chance of misrouting and unsafe prompt steering.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The activation section says the skill should be used for '任何算法竞赛题目需求', which is broader than the specialized NOI/CSP/OI framing described elsewhere. This weak boundary can make the skill activate for a wide range of contest or practice scenarios, causing inappropriate specialization and reducing predictability of system behavior.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.