ClawHub
Back to skill

Security audit

engagement-routine

Security checks for vulnerabilities and agentic risk

Overview

This is a markdown-only advisory skill for planning social-media engagement routines, with no hidden automation, install hooks, or persistence.

Installers should understand this skill provides planning guidance and may use brand profile or social strategy context to tailor advice. It does not appear to automate replies, access inboxes, collect analytics, install code, or persist anything; users should invoke it for social-media engagement routines rather than unrelated uses of the word engagement.

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 (2)

Vague Triggers

Medium
Confidence
84% confidence
Finding
The eval input "I can't keep up with comments and DMs. Help me build an engagement routine." is broad enough that a generic router could activate this skill from ordinary conversation without strong scope checks. In an agent system, over-broad activation can cause the wrong skill to run, producing off-target guidance and increasing the chance of unintended access to related context such as brand-profile/social-strategy.

Vague Triggers

Medium
Confidence
87% confidence
Finding
The input "Concretely, what should my engagement day look like?" lacks explicit boundaries tying "engagement" to social-media comment/DM workflows, so it can overlap with sales, employee, romantic, or general audience-engagement contexts. This ambiguity makes misrouting more likely, and in a multi-skill agent that can lead to irrelevant actions, incorrect assumptions about available tools, or accidental use of unrelated user context.

Static analysis

No suspicious patterns detected.