ClawHub
Back to skill

Security audit

美团特价酒店

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed Meituan hotel-deals search helper that sends hotel query parameters to a Tencent Cloud proxy and returns booking links, with some non-blocking scope and disclosure caveats.

Install only if you are comfortable with hotel search terms being sent through the publisher's Tencent Cloud proxy and with results linking out to Meituan for booking. Review ambiguous hotel requests before relying on the skill, because its trigger wording is broad and it may respond to general hotel-search intent.

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
  • Taint TrackingDirect Taint Flow, Variable-Mediated Taint Flow, Credential Exfiltration Chain
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (7)

Tainted flow: 'req' from os.environ.get (line 19, 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 · scripts/meituan_hotel.py (reported line 21)May include surrounding context.

python
headers = {"Content-Type": "application/json", "X-Proxy-Token": PROXY_TOKEN}
    req = urllib.request.Request(PROXY_URL, data=body, headers=headers, method="POST")
    try:
        with urllib.request.urlopen(req, timeout=PROXY_TIMEOUT) as resp:
            result = json.loads(resp.read().decode("utf-8"))
            return result if result is not None else {}
    except urllib.error.HTTPError as e:

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

The skill claims external real-time API access and Tencent Cloud proxying, which implies network-capable behavior, but it does not declare any explicit tool scope or allowed-tools boundary. That creates a least-privilege and governance gap: the runtime may grant broader capabilities than users or reviewers can verify, increasing the chance of unintended data access or network use.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The trigger phrases are broad generic travel terms such as '查酒店' and '订酒店', which can overlap with many ordinary user requests that may not specifically intend to use this Meituan-focused skill. Overbroad activation can cause unintended invocation, unnecessary external data sharing, and user confusion about why a third-party booking workflow was engaged.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The guidance '当用户需要搜索酒店、查找特价房型、比较酒店价格时，调用本技能' is vague and broad, so the skill may activate for many general travel-planning requests rather than only when the user wants Meituan special-offer hotel results. In context, this is moderately risky because the skill provides outbound booking links and relies on external API data, so mistaken invocation can redirect user flow and expose query data unnecessarily.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

Natural-language strings throughout the file, including the module docstring, function docstring, errors, and output text, are exclusively in Chinese. This effectively forces a specific language without any user opt-in or documented justification for a locale-specific constraint.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill transmits user-supplied search parameters and a proxy authentication token to an external hard-coded cloud endpoint without any user-facing disclosure or consent language. In the skill context, city/brand inputs are not highly sensitive by default, but undisclosed third-party transmission still creates privacy and trust risk and exposes the token to the remote service by design.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Low
Category
Not specified by scanner
Confidence
92% confidence
Finding

The skill appends train and flight search prompts that exceed the declared hotel-deals-only scope, which can steer users into adjacent actions not described in the manifest. This is not code execution or data exfiltration, but it is a scope-integrity issue that can confuse users and enable unexpected capability expansion in agent ecosystems.

Content

No source excerpt is available for this finding.