ClawHub
Back to skill

Security audit

多代理开发框架

Security checks for vulnerabilities and agentic risk

Overview

This skill is a developer workflow guide that can modify code through an agent, but its requested powers fit its stated purpose and no hidden or destructive behavior was found.

Install only if you want an agent to coordinate code changes, tests, reviews, and commits from a clear plan. Use it in a trusted repository or branch, review proposed file changes before accepting them, and do not provide a callback URL unless you understand what endpoint will receive completion details.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill declares powerful read/exec/write capabilities and repeatedly instructs the agent to implement, test, and modify code, but it does not prominently warn users that invoking the skill may execute local commands and change files. In a multi-agent orchestration workflow, this omission is especially risky because sub-agents may amplify side effects across multiple tasks before the user realizes the operational impact.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The documented callback_url parameter enables sending completion data to an external endpoint, but the skill does not warn about data leaving the local environment or what contents may be transmitted. Given the skill processes plan files, source code context, and task outputs, this can lead to unintentional disclosure of sensitive project information to third-party services.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The example trigger '使用多代理开发执行这个计划' is broad and does not constrain what repositories, files, or commands may be acted on before the skill begins orchestrating sub-agents with read/exec/write access. In this skill’s context, that ambiguity increases the chance of unintended code execution or repository modification from a casual natural-language invocation.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.