ClawHub
Back to skill

Security audit

Alpha Feed Creator

Security checks for vulnerabilities and agentic risk

Overview

This markdown-only skill describes a content collection and reporting workflow with disclosed external pushes and scheduling, but users should configure those integrations carefully.

Install only if you want an agent to collect content from configured sources, rank it, write reports, and optionally send those reports to team channels. Keep API keys and webhook tokens in environment variables or a secrets manager, review all push targets and cron schedules before enabling them, and use dry-run or test modes before recurring delivery.

SkillSpector

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 (3)

Intent-Code Divergence

Medium
Confidence
92% confidence
Finding
The documented JSON return schema describes a grading/audit result unrelated to the skill's advertised function of content collection, ranking, and group pushing. This kind of interface mismatch is dangerous because downstream agents or users may make trust decisions, trigger workflows, or suppress validation based on fabricated 'success' or 'security compliance' fields that do not reflect the actual operation performed.

Vague Triggers

Medium
Confidence
85% confidence
Finding
The description uses broad activation wording like improving efficiency, automation, batch processing, and workflow optimization, which can cause the skill to be invoked for many unrelated requests. In an agent ecosystem, this increases the chance that a high-privilege skill with exec capability is selected outside its intended scope, leading to unnecessary command execution, data access, or outbound actions.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The skill advertises automatic group pushing and scheduled execution but does not present a prominent warning near those features about external data transmission, recurring automated actions, or approval requirements. In context, this is more dangerous because the skill also declares exec capability, multi-platform collection, and enterprise/team usage, so mistaken use could repeatedly exfiltrate collected content or metadata to external channels without sufficient user awareness.

Static analysis

No suspicious patterns detected.