Back to skill

Security audit

Read Tweet

Security checks for vulnerabilities and agentic risk

Overview

This skill is a simple tweet-reading helper that uses third-party tweet proxy services, with privacy tradeoffs but no hidden install, persistence, credential access, or destructive behavior found.

Install only if you are comfortable with tweet URLs you ask it to read being sent to third-party proxy services such as fxtwitter or vxtwitter. Treat linked-article fetching or broader search as an extra step the agent should ask about first.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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 (7)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The trigger phrases are broad enough to match casual conversation, which can activate the skill when the user did not intend to send a tweet URL to a third-party proxy. In this skill, accidental activation matters because execution leads to external network requests and data transmission.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
86% confidence
Finding

The skill description and output instructions are written entirely in Chinese and direct the agent to produce structured output in Chinese, but there is no indication that the user can choose another language. This creates a language policy issue because the skill appears to impose a specific locale without opt-in.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill sends tweet URLs and request metadata to third-party proxy services without any user-facing disclosure or consent mechanism. Even if the content is public, the user's access pattern, requested tweet, and associated metadata are still being transmitted to an external service outside the primary platform.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
95% confidence
Finding

This explicitly routes user-supplied tweet URLs through a third-party proxy domain, creating an external data transmission path. The risk is elevated because the proxy is central to normal operation, so every use of the skill can disclose user-request context to an external service.

Content

Scanner excerpt · SKILL.md (reported line 29)May include surrounding context.

text
原始: https://x.com/HiTw93/status/2032091246588518683
代理: https://api.fxtwitter.com/HiTw93/status/2032091246588518683

使用 Bash 通过 curl 访问代理 URL:

External Transmission

Medium
Category
Data Exfiltration
Confidence
96% confidence
Finding

The curl command operationalizes the external transmission by directly sending the request to api.fxtwitter.com. Without safeguards, this means any matching invocation results in outbound network access to a third-party service and possible logging of the requested resource and client metadata.

Content

Scanner excerpt · SKILL.md (reported line 35)May include surrounding context.

使用 Bash 通过 curl 访问代理 URL:

bash
curl -s "https://api.fxtwitter.com/{username}/status/{tweet_id}"

返回的是 JSON 数据,从中提取推文完整内容。

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The fallback behavior expands the skill beyond the user-scoped task of reading a specific shared tweet into broader WebSearch and arbitrary external-link fetching. That scope creep can cause unintended retrieval of unrelated content, additional outbound requests, and possible exposure of user-provided data to more third parties than necessary.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
93% confidence
Finding

The fallback to a second third-party proxy increases the number of external recipients of user request data and broadens the trust boundary. This compounds privacy and integrity risk because a failure in one proxy silently shifts traffic to another external service without explicit user approval.

Content

Scanner excerpt · SKILL.md (reported line 59)May include surrounding context.

md
## 备选方案

如果 fxtwitter 不可用,依次尝试:
1. `https://api.vxtwitter.com/{username}/status/{tweet_id}`
2. 使用 WebSearch 搜索推文 ID 或正文关键词

## 注意事项

Static analysis

No suspicious patterns detected.