Vague Triggers
Medium
- Category
- Not specified by scanner
- Confidence
- 93% confidence
- Finding
该描述将“抖音分析”“抖音查询”“达人分析”等列为触发词，但未提供边界条件、排除示例或限定上下文。这些短语在日常业务对话中较常见，容易与一般性分析/查询请求重叠，从而造成技能激活范围过宽。
- Content
Security audit
Douyin Data Method
Security checks for vulnerabilities and agentic risk
Overview
This skill is a disclosed Douyin data lookup helper that uses a third-party MaxHub API key, with some scope cautions but no evidence of hidden or destructive behavior.
Install only if you intend to use MaxHub/aconfig.cn for Douyin data and are comfortable sending your API key and query data to that service. Use a scoped API key where possible, avoid providing session cookies, and make requests specific so the fuzzy routing does not call endpoints you did not intend.
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 Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
- Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
- Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)
Vague Triggers
Medium
- Category
- Not specified by scanner
- Confidence
- 89% confidence
- Finding
The routing rules allow 'fuzzy matching' and instruct the agent to call a path for uncovered requests, leaving invocation boundaries ambiguous. In a skill that transmits authenticated requests to a third-party API, ambiguous routing can cause overbroad or unintended API calls, increasing the chance of accidental data disclosure or misuse beyond the user's actual intent.
- Content
External Transmission
Medium
- Category
- Data Exfiltration
- Confidence
- 96% confidence
- Finding
The skill instructs the agent to send authenticated requests containing a bearer token to
https://www.aconfig.cn, a third-party service. Because the skill is designed to process potentially sensitive search, user, live-room, and profile data, this creates a real external transmission risk: user-provided query data and credentials are exposed to an external operator, and the skill itself acknowledges that credentials are transmitted to that service.- Content
请求格式
bash # GET curl -s "https://www.aconfig.cn/api/v1/douyin/{endpoint}?{params}" -H "$maxhub_auth_header" # POST curl -s -X POST "https://www.aconfig.cn/api/v1/douyin/{endpoint}" -H "$maxhub_auth_header" -H "Content-Type: application/json" -d '{...}'
Intent-Code Divergence
Medium
- Category
- Not specified by scanner
- Confidence
- 93% confidence
- Finding
The file's main error-handling sections state that 404/500/410 should trigger documented downgrade behavior, with 500 retried once and then downgraded, while 403 should not be retried. But the execution rules separately instruct retry on 502 and say nothing about downgrade there, creating an active contradiction in the skill's own operational guidance. This can cause the agent to follow inconsistent behavior despite the document presenting itself as a structured method.
- Content