ClawHub
Back to skill

Security audit

Backend Engineer

Security checks for vulnerabilities and agentic risk

Overview

This is a broad backend engineering guidance skill with some overbroad activation terms, but it does not install code, persist, hide behavior, or direct unrelated data access.

Install this if you want broad backend, API, cloud, DevOps, Terraform, and payment-integration guidance. Be aware it may activate on generic engineering words like deploy or AWS, so users should confirm before applying production, infrastructure, or payment-related changes.

SkillSpector

Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (5)

Vague Triggers

Medium
Confidence
95% confidence
Finding
The activation guidance is broad enough to match many ordinary backend-related requests without clear user intent to invoke this specific skill. Over-broad routing can cause unintended activation, increasing the chance that users receive domain-specific instructions or operational commands they did not explicitly request.

Vague Triggers

Medium
Confidence
97% confidence
Finding
The trigger list contains highly generic terms such as 'deploy', 'AWS', 'monitoring', and 'microservice' that are common in routine engineering conversations. This creates a prompt-routing vulnerability where the skill may activate too often, potentially overriding more appropriate tools or causing unsafe operational guidance to appear in unrelated contexts.

Shadow Command Trigger

Medium
Category
Trigger Abuse
Confidence
91% confidence
Finding
Using 'build API' as a trigger can shadow or conflict with a built-in 'build' command family, creating ambiguous routing and unexpected skill invocation. In an agent environment, command-shadowing can be abused indirectly by causing the wrong execution path to be selected for routine user input.

Shadow Command Trigger

Medium
Category
Trigger Abuse
Confidence
96% confidence
Finding
The single-word trigger 'deploy' is especially collision-prone with common built-in command names and ordinary user requests. This makes accidental or adversarial invocation easier, which is risky because the skill contains extensive infrastructure and operational guidance.

Shadow Command Trigger

Medium
Category
Trigger Abuse
Confidence
84% confidence
Finding
The trigger 'load balancer' partially overlaps with the built-in command stem 'load', which can create ambiguous matching depending on the host platform's parser. Although less direct than a standalone reserved verb, it still increases the risk of misrouting and unintended activation.

Static analysis

No suspicious patterns detected.