Back to skill

Security audit

AI Traffic Analytics

Security checks for vulnerabilities and agentic risk

Overview

The verdict is benign with medium confidence: the scan found a privacy concern about conversation excerpts in reports, while the other rated areas were acceptable or had notes.

Before installing, consider whether monthly reports may include customer conversation excerpts and who can access them. The scanner recommends anonymized paraphrases by default, with direct quotes used only after privacy review and any required authorization. The scan was partial because it could not resolve a local reference, so the evidence is incomplete.

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
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • 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 (1)

Customer conversation excerpts are included in monthly reports without a privacy warning or redaction requirement.

Medium
Category
Output Handling
Confidence
94% confidence
Finding

The skill explicitly asks for monthly quotes from sales and support conversations without requiring redaction or checking whether report recipients may access the underlying customer information. Those excerpts could disclose personal information or confidential business details to a broader audience; the privacy safeguard for raw server logs does not extend to conversation excerpts.

Content

No source excerpt is available for this finding.