Back to skill

Security audit

Lanxin App Articles

Security checks for vulnerabilities and agentic risk

Overview

The skill appears to format Lanxin article-card JSON, but its always-on broad activation and refusal-suppressing instructions make it risky for review before installation.

Install only if you want a Chinese Lanxin article-card formatter that may activate broadly. The skill should be revised to require explicit send intent, allow clarifying questions and refusals, and validate required fields and URLs before emitting JSON.

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
Findings (1)

T01 · Skill Instruction Hijacking

Error
Location
SKILL.md:20
Finding

Absolute Skill Instructions Suppress Refusals and Safety-Aware Responses

Content
View full analysis
` 或 `` 标签** ``` The final behavioral directive reinforces the same constraint at `SKILL.md:93-95`: ```markdown ## 你的行为准则 用户让你发送图文卡片 → **直接输出干净的 JSON** → 不要多余描述 → 不要格式包裹 → 完成。 ``` ### Technical Analysis The skill declares itself always active and then presents its behavioral rules as absolute requirements that must never be violated. Most of the JSON formatting restrictions are consistent with the skill's stated purpose. However, explicitly prohibiting refusal language is not necessary to implement app-article formatting and attempts to interfere with the agent's ability to reject unsafe, deceptive, unsupported, or insufficiently specified requests. The combination of global activation, high priority, mandatory output, JSON-only responses, and refusal suppression can alter the agent's safety-aware decision process whenever the skill is loaded. Although higher-level system and developer instructions should still take precedence, the skill text attempts to constrain the current session in a way characteristic of instruction hijacking. ### Attack Path 1. The skill is loaded globally because its metadata sets `openclaw.always` to `true`. 2. An attacker submits a request involving an app-article card, potentially containing a phishing destination, malicious image URL, deceptive title, or other unsafe content. 3. The skill directs the agent to emit the `appArticles` JSON payload immediately. 4. The refusal prohibition discourages ...[truncated 1051 chars]
Remediation
View remediation
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (6)

Vague Triggers

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The trigger list includes very generic terms such as “图文” and “文章”, which are likely to appear in ordinary user conversation and can cause the skill to activate unintentionally. In this skill, accidental activation is more dangerous because the body then forces a rigid JSON-only response, potentially hijacking normal interactions and causing unintended outbound message formatting or delivery behavior.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill instructions force a single output mode—raw JSON only, no explanation, no refusal language, and no alternate formatting—without user opt-in. This can override safer assistant behavior, suppress clarifications, and make it easier for the skill to coerce the agent into structured message emission even when the request is ambiguous or unsafe.

Content

No source excerpt is available for this finding.

Overly Broad Trigger

Low
Category
Trigger Abuse
Confidence
91% confidence
Finding

The trigger “蓝信” is too short and broad for a high-priority skill, so any mention of Lanxin could invoke the skill even when the user is not asking to send an appArticles card. Because the skill then mandates a JSON-only response, this broad trigger increases the chance of unintended tool-routing and response corruption.

Content

No source excerpt is available for this finding.

Overly Broad Trigger

Low
Category
Trigger Abuse
Confidence
98% confidence
Finding

The trigger “图文” is extremely generic and likely to match many unrelated conversations about images, documents, or articles. In this context, that makes the skill materially riskier because accidental invocation can force raw JSON output and bypass normal conversational handling.

Content

No source excerpt is available for this finding.

Overly Broad Trigger

Low
Category
Trigger Abuse
Confidence
97% confidence
Finding

The trigger “文章” is a common everyday term and is not specific enough to indicate intent to send a Lanxin article card. This creates a realistic risk of false activation, which is amplified by the skill’s strict requirement to output only message JSON.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

The description string is written only in Chinese ('蓝信图文卡片发送技能'), which indicates a fixed language choice in user-facing metadata. This can violate language/locale policy when the skill does not offer any user language selection or document that it is intentionally region-specific.

Content

No source excerpt is available for this finding.