Back to skill

Security audit

投资性房地产·逐条挂源

Security checks for vulnerabilities and agentic risk

Overview

ClawScan marked this skill suspicious because scanners report undeclared remote tool discovery, privacy-sensitive document uploads without a clear consent warning, and an undisclosed local usage marker; scan coverage is partial.

Consider the suspicious verdict and the incomplete scan before installing. In particular, review whether you are comfortable sending report attachments to an external parsing service, with the recipient and consent requirements unclear, and with a local usage marker being created. The checker may also fail to catch an incorrect transfer amount, and the skill's declared tool scope is absent.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (8)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

The supplied code performs remote service discovery and tool enumeration rather than processing financial disclosures. Nothing in the chunk connects the catalog or routing endpoint to investment-property extraction, so these operations are not established as supporting implementation details of the declared purpose. The code does not show report parsing, extraction, quotation, or citation generation. It also does not show document exfiltration or execution of the discovered tools.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

The reported daily usage marker is an undisclosed persistent side effect unrelated to validating financial disclosures, including during demo and self-test execution. It leaves a local record of skill usage outside the workspace, but the evidence does not show transmission, document capture, or a high-impact tracking mechanism.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · README.md (reported line 53)May include surrounding context.

CATALOG="https://cuecue.cn/api/mcp-catalog" curl -s "$CATALOG" | head -c 2000

从目录里取本组的路由地址,确认工具在场

curl -s -X POST "$ROUTING_URL"
-H 'Content-Type: application/json'
-d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'

text

The workflow directs user-provided documents to a parsing service without warning about third-party data disclosure.

Medium
Category
Output Handling
Confidence
90% confidence
Finding

The workflow routes user-provided attachments through an external parsing service without explicitly explaining that document contents leave the user's environment or requiring permission. Published financial reports reduce the confidentiality risk, but user-supplied files may contain unpublished information or annotations.

Content

No source excerpt is available for this finding.

Skill declares no tool scope ('permissions' or 'allowed-tools') but code capabilities were detected: file_read, file_write.

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

The workflow sends financial-report attachments to an external parsing service without warning about disclosure of document contents.

Medium
Category
Output Handling
Confidence
94% confidence
Finding

The instructions allow attachment parsing to begin when evidence resides in a PDF, but do not require approval for disclosure of its contents to an external service. References to service availability and billing do not establish informed authorization to upload confidential user-provided material.

Content

No source excerpt is available for this finding.

The transfer checker claims to verify quoted amounts but checks only amount presence and destination wording.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The transfer checker detects numeric presence and compares destination keywords, but never compares the asserted transfer amount with the quoted amount. An incorrect or deliberately altered amount can therefore pass this check, undermining the financial register's claimed evidence-backed integrity.

Content

No source excerpt is available for this finding.

The checker persistently records usage in the user's home directory, unrelated to financial-disclosure validation.

Low
Category
Excessive Agency
Confidence
99% confidence
Finding

mark_first_run creates a dated marker under ~/.cue-usage and silently suppresses errors, leaving persistent usage evidence outside the requested input and output locations. The marker contains no financial contents and is not transmitted, which limits the issue to an unnecessary local privacy and filesystem side effect.

Content

No source excerpt is available for this finding.