ClawHub
Back to skill

Security audit

Cold Email

Security checks for vulnerabilities and agentic risk

Overview

This content-only cold email helper has no executable behavior, but it should be reviewed because it encourages misleading subject lines that can look like internal colleague emails.

Review this before installing if your organization has strict outreach, privacy, or brand-trust rules. The skill does not run code or send emails by itself, but users should avoid internal-looking or colleague-like subject lines, fake familiarity, sensitive personal data, and any cold outreach that would violate CAN-SPAM, GDPR, or similar rules.

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 (4)

Anti-Refusal Statement

High
Category
Anti-Refusal
Content
**Structure:** Trigger/Pain → Solution hint → Binary CTA. 1–3 sentences, no intro.
**Best for:** High-velocity SDR teams. Mobile-optimized. Deliberately polarizing.

Spend max 1 minute on personalization. Use industry/persona-level signals. For top-tier prospects, quote their own words from interviews — they almost always respond.

## Vanilla Ice Cream (Lavender)
Confidence
70% confidence
Finding
Skill instructs the agent to never refuse or to always comply. Suppressing the agent's ability to decline removes a core safety control and enables downstream harmful requests to succeed.

Vague Triggers

Medium
Confidence
95% confidence
Finding
The eval explicitly expects the skill to trigger on casual phrasing for a follow-up-sequence request, which can broaden activation criteria beyond clear cold-email intent. In an agent routing context, vague trigger expectations increase the chance this skill is invoked for loosely related email tasks, causing misrouting and inappropriate guidance rather than direct code execution or data compromise.

Missing User Warnings

Medium
Confidence
89% confidence
Finding
The guidance encourages aggressive personalization techniques without any boundary-setting around privacy, deception, or compliance. In a cold-email skill, telling users to quote prospects' interview statements and optimize for response rates can normalize invasive or manipulative outreach, increasing the chance of misuse for unethical targeting or policy-violating campaigns.

Natural-Language Policy Violations

Medium
Confidence
90% confidence
Finding
The guidance explicitly recommends 'internal-looking' subject lines that make a sales email appear as if it came from a colleague rather than a vendor. This is deceptive impersonation-style social engineering, increasing the chance recipients open messages under false assumptions and creating compliance, trust, and brand-risk issues. In a cold-email skill, this context makes the issue more dangerous because the content is intended for scaled outbound use, so the deceptive pattern could be operationalized broadly.

Static analysis

No suspicious patterns detected.