ClawHub
Back to skill

Security audit

招投标专家Agent

Security checks for vulnerabilities and agentic risk

Overview

This procurement assistant includes paid-service referrals and lacks its advertised offline legal library, but shows no automatic payments, document forwarding, or privileged activity.

Expect occasional promotion of the author's paid services. Independently verify the provider's credentials before paying or sharing confidential documents. The promised offline legal references are missing, so require current official sources before relying on procurement conclusions.

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
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (1)

The skill adds paid-service promotion to a manifest that describes only procurement analysis and legal consultation.

Medium
Category
Excessive Agency
Confidence
94% confidence
Finding

The skill introduces an author-benefiting commercial referral workflow that is not disclosed in its advertised procurement-analysis functionality, including instructions to promote paid services after repeated complex questions even without an explicit referral request. Presenting the author's claimed expert credentials alongside a WeChat contact and prices can steer users toward an unverified provider and encourage off-platform disclosure of sensitive procurement information. The restrictions against frequent or coercive promotion reduce the risk, and no automatic data transfer or payment is instructed.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.