Back to skill

Security audit

Solo CEO

Security checks for vulnerabilities and agentic risk

Overview

The skill appears intended for task coordination, but it asks agents to spawn worker agents and persist plans and memory in ways that could retain sensitive task or customer context without enough boundaries.

Review this skill before installing if your tasks include customer data, confidential project details, or regulated information. Use it only where persistent plan files and worker MEMORY.md files are acceptable, and periodically inspect or clear those files. The artifact does not show malicious payloads, but its memory and delegation model needs stronger user control.

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
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (6)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The entire skill is written as a Chinese-only operating specification and does not indicate that users may choose another language. This can violate language or locale policy when a skill implicitly forces a specific language without opt-in or clear regional justification.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The skill mixes a rule that task-specific context must not be persisted with repeated instructions to update MEMORY.md after tasks with newly learned information. In practice, agents may misclassify task details, customer data, or sensitive operational context as a 'preference' and write it to persistent storage, causing unauthorized retention and cross-task data leakage.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The instructions tell agents to read and modify MEMORY.md as part of normal execution, but do not warn the user that the skill will perform persistent file access and writes. This creates a transparency and consent issue and can silently accumulate sensitive preferences or task-derived data across sessions.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The planning phase instructs the agent to create plan.md in the workspace containing task names, goals, assignments, and progress, without telling the user that their task data will be written to disk. This can persist potentially sensitive business or personal information beyond the active session and expose it to later runs or other workspace readers.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

This markdown template mandates Chinese-language instructions and examples throughout, starting from the title and introductory text. Under the policy, forcing a specific language without offering the user a language or locale choice is a natural-language policy violation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
82% confidence
Finding

This markdown template is primarily written in Chinese while also prescribing several section headings and formats in English, such as 'Task', 'Project', 'Executive Summary', and 'Quick Task'. Because the template does not state that language should follow user preference or allow locale selection, it may impose a language/locale choice on downstream skill usage.

Content

No source excerpt is available for this finding.