Back to skill

Security audit

Aivi Engagement

Security checks for vulnerabilities and agentic risk

Overview

This skill is coherent and not deceptive, but it can launch paid AI voice and SMS outreach to real people without clearly documented confirmation, consent, or compliance controls.

Review this before installing if your agents may contact customers or prospects. Only connect it to accounts where users are authorized to place calls/texts, have consent records, and can manage TCPA, HIPAA, FDCPA, opt-out, billing, and audit requirements. Revoke the OAuth connector when it is no longer needed.

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
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill promotes launch_sequence as a routine action but does not clearly warn that it triggers outbound AI voice and SMS contact to third-party phone numbers. In an agent context, that omission can cause users or downstream agents to initiate real-world outreach without adequate consent, compliance, or user awareness, increasing legal, privacy, and reputational risk.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The phrase "in one command" is overly broad and suggests the skill may be invoked for multiple high-impact actions without clearly scoped user intent or guardrails. In this skill's context—lead scoring, enrichment, and launching voice/SMS outreach across sensitive sectors like debt collection and healthcare—ambiguous invocation language increases the risk of unintended automation, compliance violations, or accidental triggering of outreach workflows.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
7% confidence
Finding

I do not see a clear language or locale restriction in the manifest text. Although the description references multiple platforms, it does not force a specific language or locale, so this should not be treated as a substantive policy violation.

Content

No source excerpt is available for this finding.