ClawHub
Back to skill

Security audit

API 探针

Security checks for vulnerabilities and agentic risk

Overview

This API testing skill is coherent, but it includes live destructive, security, and load-testing examples without strong authorization or safety gates.

Install only if you will use it on APIs you own or are explicitly authorized to test. Before running generated tests, restrict targets to staging or disposable environments, remove destructive payloads by default, set timeouts and rate limits, avoid real credentials, and require separate confirmation for security or load testing.

A.I.G

Vulnerability Patterns
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (1)

T09 · Insecure Skill Coding Practices

Error
Location
SKILL.md:561
Finding

Active Destructive and High-Volume API Testing Lacks Mandatory Authorization Safeguards

Content
View full analysis
0.99 assert qps > 100 ``` **SQL injection test (`SKILL.md:604-617`):** ```python def test_sql_injection(self): payloads = [ "' OR '1'='1", "'; DROP TABLE users; --", "1 UNION SELECT * FROM users" ] for payload in payloads: response = requests.get( f"{self.base_url}/api/users", params={"search": payload} ) assert response.status_code in [400, 403, 500] ``` **Rate-limit test (`SKILL.md:645-651`):** ```python def test_rate_limiting(self): responses = [] for _ in range(150): response = requests.get(f"{self.base_url}/api/users") responses.append(response.status_code) assert 429 in responses ``` ### Technical Analysis Security and performance testing legitimately require network access, but the examples do not enforce least-privilege safeguards around that access. In particular: - The SQL test includes a payload con ...[truncated 2743 chars]
Remediation
View remediation

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (8)

External Script Fetching

High
Category
Supply Chain
Confidence
90% confidence
Finding

Remote code is downloaded and executed. This bypasses code review and could introduce malicious code.

Content

Scanner excerpt · SKILL.md (reported line 30)May include surrounding context.

md
| Swagger/OpenAPI 文档 | JSON/YAML 格式 | 解析接口定义，生成测试用例 |
| 接口文档 | Markdown/HTML | 提取接口信息，构建测试 |
| 接口地址 | URL + 方法 | 直接发起请求测试 |
| cURL 命令 | shell 命令 | 解析并转换为测试脚本 |
| Postman Collection | JSON 导出 | 导入并执行测试 |

### 2. 接口测试框架

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill advertises very broad trigger conditions such as generic mentions of API testing, Swagger, Postman, and interface validation, which can cause it to activate in contexts the user did not clearly intend. Because this skill contains capabilities for live request execution, mock services, performance testing, and security testing, overbroad activation increases the chance of unsafe or unauthorized actions being suggested or initiated against real systems.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The description includes high-impact behaviors like performance testing, security scanning, destructive HTTP methods, and direct request execution without an explicit warning to avoid production or unauthorized targets. In context, this makes the skill more dangerous because it is specifically designed to generate or run actions that can modify data, exhaust resources, or probe defenses on live systems.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
90% confidence
Finding

This sample code performs a live authentication request using embedded test credentials and then reuses the returned bearer token for further API access. In an agent skill context, generating or encouraging direct credentialed requests to arbitrary user-supplied base URLs can lead to credential misuse, unauthorized access attempts, and transmission of secrets to attacker-controlled endpoints.

Content

Scanner excerpt · SKILL.md (reported line 72)May include surrounding context.

md
def get_token(self):
        """获取认证 Token"""
        response = requests.post(f"{self.base_url}/auth/login", json={
            "username": "testuser",
            "password": "Test1234"
        })

External Transmission

Medium
Category
Data Exfiltration
Confidence
70% 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 · SKILL.md (reported line 236)May include surrounding context.

md
}
        }
        """
        response = requests.post(self.graphql_url, json={"query": query}, headers=self.headers)
        
        assert response.status_code == 200
        data = response.json()

External Transmission

Medium
Category
Data Exfiltration
Confidence
70% 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 · SKILL.md (reported line 262)May include surrounding context.

md
}
        }
        
        response = requests.post(
            self.graphql_url,
            json={"query": mutation, "variables": variables},
            headers=self.headers

External Transmission

Medium
Category
Data Exfiltration
Confidence
70% 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 · SKILL.md (reported line 624)May include surrounding context.

md
]
        
        for payload in payloads:
            response = requests.post(
                f"{self.base_url}/api/users",
                json={"username": payload, "email": "test@example.com"}
            )

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
89% confidence
Finding

The manifest description and main instructional content are presented entirely in Chinese, with no indication that the user may choose another language. This can violate a language/locale policy when the skill effectively assumes a fixed language rather than offering a user preference.

Content

No source excerpt is available for this finding.