Back to skill

Security audit

hsciq-mcp

Security checks for vulnerabilities and agentic risk

Overview

ClawScan rates this skill suspicious with medium confidence, chiefly because its instructions reportedly allow product details and images to be submitted for external human review without explicit consent, and trigger on some generic phrases.

Consider avoiding installation until you can verify that the skill only activates in relevant customs or trade contexts and requires your explicit approval before sending product details or images for external review. The scan is incomplete: it covered 80% of components, and interpretation of part of the JavaScript was degraded, so the findings do not establish that unreported behavior is safe.

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
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Taint TrackingDirect Taint Flow, Variable-Mediated Taint Flow, Credential Exfiltration Chain
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
Findings (6)

Tainted flow: 'req' from os.getenv (line 95, credential/environment) → urllib.request.urlopen (network output)

Critical
Category
Data Flow
Confidence
90% confidence
Finding

Credentials or environment variables flow to a network sink. This is a high-confidence indicator of credential exfiltration.

Content

Scanner excerpt · hsciq_client.py (reported line 98)May include surrounding context.

python
req = urllib.request.Request(url, data=data, headers=headers, method='POST')

        try:
            with urllib.request.urlopen(req, timeout=30) as response:
                body = response.read().decode('utf-8')
        except urllib.error.HTTPError as e:
            error_body = e.read().decode('utf-8') if e.fp else ""

MCP Config Access

High
Category
Agent Snooping
Confidence
80% confidence
Finding

Skill accesses MCP server configuration files (mcp.json). MCP configs contain server URLs, authentication tokens, and tool definitions — reading them allows the skill to discover and potentially abuse other tool integrations.

Content

Scanner excerpt · hsciq-client.js (reported line 314)May include surrounding context.

js
console.log('  list-ocean-carriers   (list enabled ocean carriers with SCAC codes)');
        console.log('  search-ocean-carrier-notice [--keywords <kw>] [--carrierCode <scac>] (last 30 days only)');
        console.log('  get-ocean-carrier-notice-detail --noticeId <guid> (last 30 days only)');
        console.log('  list-tools            (list available MCP tools via tools/list)');
        process.exit(0);
    }
    console.log(JSON.stringify(result, null, 2));

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

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.

Automatic triggers include generic phrases unrelated to customs classification.

Medium
Category
Trigger Abuse
Confidence
98% confidence
Finding

The skill automatically activates on phrases such as “提交审核” (submit for review), “专家确认” (expert confirmation), and “分类列表” (category list), without requiring customs or trade context. These phrases commonly occur in unrelated workflows, and no exclusion conditions constrain their use.

Content

No source excerpt is available for this finding.

Automatic expert-review submission lacks a user-facing privacy warning.

Medium
Category
Output Handling
Confidence
96% confidence
Finding

The example directs automatic submission of product information and images when a user doubts an AI result, without requiring a user-facing disclosure before upload. Although human review is documented elsewhere, that does not ensure the affected user understands that potentially confidential product material will leave the conversation and be viewed by external reviewers.

Content

No source excerpt is available for this finding.

User uncertainty is treated as authorization to disclose product information and images to external human reviewers.

Medium
Category
Data Exfiltration
Confidence
98% confidence
Finding

The workflow treats uncertainty about an AI classification as permission to create an external consultation and upload product details and images. A request to check accuracy is not authorization to disclose those materials, so following this example can expose proprietary information without informed consent.

Content

No source excerpt is available for this finding.