ClawHub
Back to skill

Security audit

Outreach Scout

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed outreach helper that searches public platforms, drafts replies for user approval, and keeps a local activity log, with automation and privacy caveats users should configure carefully.

Install only if you want an agent-assisted outreach workflow. Keep posting set to approval-required, avoid storing unnecessary personal data, check each platform's rules before engaging, and add a retention policy for the local scout log.

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
  • 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 (3)

Vague Triggers

Medium
Confidence
95% confidence
Finding
The skill description includes broad activation cues like marketing, lead generation, audience building, and growing product awareness, which can match many ordinary user requests and cause the skill to trigger outside a narrowly intended outreach context. Over-broad activation increases the chance of unintended execution of a growth/outreach workflow that performs monitoring, reply drafting, and activity tracking without the user explicitly asking for this specific behavior.

Vague Triggers

Medium
Confidence
94% confidence
Finding
The manual trigger condition, 'When asked to scout, find leads, or do outreach,' is ambiguous and permissive, so ordinary requests about research, community participation, or marketing advice could invoke a workflow that searches public platforms, profiles potential leads, and drafts promotional responses. Because the workflow also reads config, checks logs, and tracks interactions, accidental activation can lead to unreviewed collection and preparation of outreach actions beyond user intent.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The tracking section instructs the agent to log thread URLs, timestamps, actions taken, engagement results, and conversions, but provides no privacy, retention, consent, or data-minimization guidance. In context, this skill monitors individuals on social platforms and records engagement history for lead generation, so missing safeguards can create privacy, compliance, and reputational risks if personal data or identifiable behavioral data is stored indiscriminately.

Static analysis

No suspicious patterns detected.