ClawHub
Back to skill

Security audit

Api Magic Gen

Security checks for vulnerabilities and agentic risk

Overview

This skill appears to be an API-generation helper, but its documentation mixes in unrelated review/scoring behavior and grants broad read, write, and command-execution capabilities without clear limits.

Review this skill before installing. Use it only in a limited development workspace, avoid granting production credentials or broad filesystem access, and require explicit approval before any command execution, file writes, deployment changes, or external API calls.

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 (4)

Intent-Code Divergence

Medium
Confidence
95% confidence
Finding
The documented return schema describes a code-review/scoring tool, while the rest of the skill presents an API-generation and magic-api operational capability. This mismatch can cause the agent or user to invoke the skill under false assumptions, leading to unsafe automation decisions, incorrect handling of outputs, or misuse of elevated capabilities such as exec/write based on deceptive documentation.

Intent-Code Divergence

Medium
Confidence
95% confidence
Finding
The parameter section frames the skill as processing review/audit content rather than generating or managing APIs, directly conflicting with the broader skill description. Such semantic inconsistency is dangerous because policy engines, orchestrators, or users may authorize the skill for a lower-risk task while it actually advertises higher-risk capabilities like command execution and file operations.

Vague Triggers

Medium
Confidence
90% confidence
Finding
The activation text uses broad triggers such as general API integration, webhook configuration, system connection, monitoring, logging, and deployment scenarios without tight scope boundaries. Overbroad invocation conditions increase the chance the agent will select this skill in unrelated or sensitive contexts, unintentionally granting it access to powerful behaviors and expanding attack surface.

Missing User Warnings

Medium
Confidence
97% confidence
Finding
The skill advertises file processing, external API calls, and system command execution, yet does not provide a clear upfront warning or consent model about the potential system impact. In the context of an agent skill, this is especially dangerous because users may invoke it expecting documentation/generation help, while the skill’s available tools can modify files, run commands, and interact with external systems.

Static analysis

No suspicious patterns detected.