ClawHub
Back to skill

Security audit

magpie

Security checks for vulnerabilities and agentic risk

Overview

This skill is a local Chinese A-share stock monitor that documents localhost API calls, with disclosed watchlist and alert changes and no bundled executable or external exfiltration.

Install only if you run and trust the local magpie daemon, because the skill can read your local stock watchlist and alert history and can add or remove local watchlist or alert rules when you ask. Review ambiguous prompts before letting it create alerts.

A.I.G

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

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

External Transmission

Medium
Category
Data Exfiltration
Content
### 龙虎榜 lhb

```bash
curl -s "http://127.0.0.1:17891/api/v1/lhb"                      # today
curl -s "http://127.0.0.1:17891/api/v1/lhb?date=2026-05-13"      # specific date
```
Confidence
60% confidence
Finding
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Natural-Language Policy Violations

Medium
Confidence
90% confidence
Finding
The output-style section mandates Chinese phrasing for disclaimers and formatting, and the skill overall is written to respond in Chinese without stating that this is optional or user-selectable. Under the policy, forcing a specific language without user opt-in is a natural-language locale violation unless clearly justified as a region-specific tool.

Vague Triggers

Medium
Confidence
95% confidence
Finding
The trigger list includes broad generic phrases such as '提醒我', '收盘', and common stock names, which can cause the skill to activate in contexts the user did not intend. Over-broad activation increases the chance that user requests are routed to a local daemon and interpreted as stock-monitoring actions, creating unintended data access or side effects such as alert creation.

Static analysis

No suspicious patterns detected.