Back to skill

Security audit

信贷尽调

Security checks for vulnerabilities and agentic risk

Overview

The skill mostly behaves like a disclosed Cue research launcher, but it mixes credit-diligence work with high-net-worth prospect mining and runs/upgrades an external runner from the user’s home directory.

Install only if you are comfortable with a Chinese-language Cue workflow that uses your Cue account, consumes credits after confirmation, clones/updates and runs the external cue-skills runner, and may expose broader prospecting features beyond credit diligence. Review the upstream runner source and avoid using the prospect-mining workflow unless it fits your compliance and privacy obligations.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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 (3)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The manifest description and the full user-facing instructions are written entirely in Chinese, including prompts the agent is instructed to present to the user. There is no indication that language can be adapted to user preference or that the Chinese-only behavior is a justified locale-specific constraint.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The manifest frames this skill as a "信贷尽调" tool for corporate credit due diligence, pre-lending review, and risk analysis. However, L23 describes extracting equity-incentive and placement participant lists to identify "潜在高净值客群" for wealth-management/business-development use, which is a distinct customer-prospecting capability not justified by the stated credit-diligence purpose.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Low
Category
Not specified by scanner
Confidence
81% confidence
Finding

The manifest describes a credit-diligence research workflow centered on enterprise fundamentals, lending pre-due-diligence, and risk control. L22 adds a "产业集团战略全景分析" capability for dissecting conglomerates' industry-chain layout and global policy risk for cross-border benchmarking, which is broader strategic/competitive intelligence rather than an obvious requirement of credit due diligence.

Content

No source excerpt is available for this finding.