ClawHub
Back to skill

Security audit

Agent Browser

Security checks for vulnerabilities and agentic risk

Overview

This browser automation skill is mostly coherent, but it enables broad browser control, login/form automation, script execution, scraping, and scheduled monitoring without enough scoping detail.

Review before installing. Use this skill only when you want an agent to control a browser, and require explicit confirmation for logins, form submissions, purchases, account changes, script execution, scraping authenticated pages, and scheduled monitoring. Check where screenshots, scraped data, and config files are stored and delete them when no longer needed.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • YARA SignaturesMalware Match, Webshell Match, Cryptominer Match
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

YARA rule 'agent_skill_mcp_tool_poisoning_metadata': MCP/tool metadata poisoning indicators in tool schemas or skill manifests [agent_skills]

High
Category
YARA Match
Content
---
AIGC:
    ContentProducer: Minimax Agent AI
    ContentPropagator: Minimax Agent AI
    Label: AIGC
    ProduceID: "00000000000000000000000000000000"
    PropagateID: "00000000000000000000000000000000"
    ReservedCode1: 304502205c02c1ea86957276d0a274b909a4e0067db9ea8c5ca21de30747310b6f3dfabb022100cc035501daaa6e642da4c69ca11f72efa6fb970a6f5e26f711ebb9e3bf1ed3a9
    ReservedCode2: 3045022055ff431f1fac57a338871581c0d1ae5acd1b4c945ca1e40c60062b60cae4b531022100dd7076f1dbf9ed9d9cff03cb5b993b4f86f4988d6f289570a12a6eba8ef20409
description: 浏览器智能控制。自动化操作、截图、填表、数据抓取。
metadata:
    category: 生产力
    emoji: "\U0001F310"
    triggers:
        - 浏览器
        - browser
        - 自动化
        - 截图
        - 填表
        - 抓取
name: agent-browser
---
Confidence
80% confidence
Finding
YARA rule matched a hack tool or exploit indicator (offensive tools, reconnaissance, privilege escalation, or exploit frameworks).

Vague Triggers

Medium
Confidence
96% confidence
Finding
The trigger list is very broad and includes generic terms like 'browser', 'automation', 'screenshot', and 'scrape' that commonly appear in ordinary user requests. This can cause the skill to activate outside a narrowly intended context, increasing the chance that powerful browser capabilities are invoked unintentionally or in response to ambiguous prompts.

Vague Triggers

Medium
Confidence
91% confidence
Finding
The usage guidance says users can 'directly tell me' broad actions like opening pages, taking screenshots, and filling forms, but it does not define when the skill should or should not take over. In a skill with browser control, scraping, login support, and script execution, vague invocation rules raise the risk of over-broad activation and unintended sensitive actions.

Static analysis

No suspicious patterns detected.