Back to skill

Security audit

Agentsop Crewai

Security checks for vulnerabilities and agentic risk

Overview

This is a documentation-only CrewAI guidance skill with no executable payload, install script, hidden data access, or automatic persistence.

Before installing, check whether you are comfortable with a CrewAI SOP that is largely Chinese-language and may activate on broad multi-agent collaboration requests. It does not appear to install code or access data by itself.

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
  • 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)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill content is predominantly written in Chinese and presents its operating guidance in that language without offering any user language negotiation or documenting a locale restriction. In practice, this can cause the agent to respond in an unexpected language, making safety-critical setup details, framework caveats, and operational constraints harder for users or downstream reviewers to understand and validate.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The trigger for OP-1 is broad enough to activate on many ordinary requests that merely mention collaboration, roles, or CrewAI, even when the user is not asking for this specific skill. In an agent-routing context, overbroad activation can cause incorrect framework selection, unintended workflow injection, and propagation of inappropriate multi-agent guidance into unrelated tasks.

Content

No source excerpt is available for this finding.