Back to skill

Security audit

Ouyi API Tool

Security checks for vulnerabilities and agentic risk

Overview

This skill is a plausible Ouyi API bridge, but it can send broad user prompts to a third-party service without a clear opt-in or sensitivity check.

Review before installing. Use this only when you intentionally want prompts processed by the ouyi API, and avoid sending secrets, personal data, regulated data, or proprietary code unless you trust that provider and have appropriate approval. Prefer adding an explicit confirmation step and narrower activation rules before deployment.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (17)

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
85% confidence
Finding

The skill instructs the agent to read and execute a local Node.js script and references API-key handling, but it does not declare any explicit tool scope or allowed tools. That omission weakens least-privilege controls and can let the skill be invoked with broader capabilities than users or hosts expect, increasing the chance of unintended code execution or environment access.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The activation text is overly broad, covering generic investing, technical questions, and other AI-reasoning tasks without clear boundaries. This can cause the skill to trigger in many unrelated contexts and route arbitrary user data into an external script/API flow, expanding exposure and making misuse or accidental data exfiltration more likely.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The documentation tells operators to send user prompts to a third-party API but does not clearly disclose that user content leaves the local environment or describe privacy/security implications. Users and integrators may unknowingly transmit sensitive prompts, credentials, or proprietary data to an external service without informed consent or handling guidance.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
78% confidence
Finding

The skill is explicitly configured to send prompt content to an external endpoint, which is a real data-transfer risk in the absence of safeguards. In this context the danger is increased because the skill is broad in scope and meant for general reasoning tasks, so highly sensitive user content could be forwarded off-platform unexpectedly.

Content

Scanner excerpt · SKILL.md (reported line 26)May include surrounding context.

md
- **API Key**: 用户需要用自己的 ouyi API key 替换脚本中的 `API_KEY`
- **模型**: `ouyi-openclaw`(内部映射到 gpt-5-nano-2025-08-07)
- **Endpoint**: `https://api.rcouyi.com/v1/chat/completions`

## 使用方式

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The description is written only in Chinese and the tags include 'chinese,' which signals a language-specific constraint without indicating user choice or opt-in. Under the policy, forcing or assuming a specific language without offering choice is a natural-language policy concern.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The skill advertises very broad activation criteria such as general 'AI reasoning', investment, gold analysis, and technical questions, which overlap with many ordinary user requests. This increases the chance the agent invokes the skill unexpectedly and routes user prompts to a third-party service without clear necessity or informed consent.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The documentation instructs the agent to run a local script and forward user input to an external API, but it does not require any user-facing warning that prompts may leave the local environment. This creates a real data handling risk because sensitive prompts could be transmitted to a third party without transparency, consent, or minimization controls.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
84% confidence
Finding

The skill explicitly specifies a third-party endpoint for chat completions, confirming that user prompts are transmitted off-platform. In this context, the endpoint itself is not inherently malicious, but the external transmission is security-relevant because the skill is designed to send arbitrary user content to a remote service and the surrounding documentation lacks meaningful safeguards.

Content

Scanner excerpt · ouyi/SKILL.md (reported line 26)May include surrounding context.

md
- **API Key**: 用户需要用自己的 ouyi API key 替换脚本中的 `API_KEY`
- **模型**: `ouyi-openclaw`(内部映射到 gpt-5-nano-2025-08-07)
- **Endpoint**: `https://api.rcouyi.com/v1/chat/completions`

## 使用方式

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill advertises very broad activation triggers such as general investment, technical questions, and requests for AI reasoning, which can cause it to activate for many ordinary prompts outside a narrowly intended scope. Over-broad routing increases the chance that users are unintentionally sent through this skill and have their prompts processed by an external service without clear necessity or informed consent.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The documentation instructs the agent to send user prompts to a third-party API but does not require an explicit warning or consent flow before transmitting potentially sensitive user content. This creates a privacy and data-governance risk because users may reasonably believe they are interacting locally while their input is actually being forwarded externally.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
89% confidence
Finding

The skill is explicitly configured to transmit prompts to an external endpoint, which is not inherently malicious but is a real security and privacy concern in the absence of safeguards. In this skill's context, the danger is increased because the skill is designed for broad conversational analysis, so arbitrary user input may be exfiltrated to a remote service.

Content

Scanner excerpt · ouyi/ouyi/SKILL.md (reported line 26)May include surrounding context.

md
- **API Key**: 用户需要用自己的 ouyi API key 替换脚本中的 `API_KEY`
- **模型**: `ouyi-openclaw`(内部映射到 gpt-5-nano-2025-08-07)
- **Endpoint**: `https://api.rcouyi.com/v1/chat/completions`

## 使用方式

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The description and tags are explicitly Chinese-oriented, which can bias routing toward Chinese-language behavior without confirming the user's language preference. While not inherently dangerous by itself, this can lead to consent and usability issues, and in combination with external API use may send user content to a provider selected by language heuristics rather than explicit opt-in.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The script sends raw user-supplied prompts to a third-party service at api.rcouyi.com and authenticates with an API key, but provides no disclosure, consent flow, redaction, or policy guardrails before transmitting potentially sensitive user data. In this skill context, users may submit investment, technical, or analytical queries that can contain confidential business, personal, or regulated information, so silent exfiltration to an external provider creates a real privacy and compliance risk.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The script forwards arbitrary user-supplied prompt content directly to a third-party API endpoint without any built-in disclosure, consent check, or filtering. This creates a real data-exposure risk because users may provide sensitive business, personal, or credential-like content that is then transmitted off-platform, and the skill metadata explicitly instructs operators to use this script for user requests, making the external transfer part of normal execution.

Content

No source excerpt is available for this finding.

Vague Triggers

Low
Category
Not specified by scanner
Confidence
84% confidence
Finding

This manifest description says the tool 'supports direct conversational analysis,' but it does not specify what phrases, contexts, or boundaries should trigger the skill. In a manifest file, such broad wording can make activation scope unclear and increase the chance of unintended invocation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
91% confidence
Finding

The description states the skill is an "ouyi API 调用工具" and tags it as "chinese," which implies a language-specific skill experience. For manifest text, this can be a natural-language policy concern when a specific language is effectively imposed without documenting user choice or a clear region-specific justification.

Content

No source excerpt is available for this finding.

Vague Triggers

Low
Category
Not specified by scanner
Confidence
88% confidence
Finding

The metadata description is broad enough to trigger this skill for generic 'analysis' tasks, including investment and technical questions, without tightly constraining when it should be used. In an agentic environment, overly broad invocation criteria can cause the system to route unrelated or sensitive prompts to a skill that executes a local script and calls an external API, increasing the chance of unintended data exposure or unsafe tool use.

Content

No source excerpt is available for this finding.