Back to skill

Security audit

Ask Church: Consciousness, Philosophy and Soul Q&A

Security checks for vulnerabilities and agentic risk

Overview

The skill is rated suspicious because scanners report public posting of questions and names, plus a daily scheduled task, beyond its stated on-demand answer-retrieval purpose.

Consider whether you are comfortable with questions and optional names being made public and with recurring scheduled requests. Skillspector’s scan was partial: it inspected the single reported Markdown component but could not resolve several local path-like references, so the evidence is incomplete. A separate scan reported no findings, but that does not establish that the skill is safe.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (2)

The Q&A workflow publishes questions and optional names as public conversations, beyond the manifest's stated answer-retrieval purpose.

Medium
Category
Excessive Agency
Confidence
96% confidence
Finding

The default Q&A endpoint publicly publishes submitted questions and optional names, potentially exposing sensitive personal or conversational information when users expect only answer retrieval. The explicit publication warning reduces the risk and argues against concealed exfiltration, but the skill does not require the agent to obtain informed consent before submitting content.

Content

No source excerpt is available for this finding.

The skill instructs the agent to create a persistent daily cron job despite being described as an on-demand Q&A tool.

Medium
Category
Excessive Agency
Confidence
97% confidence
Finding

The daily-practice section directs the agent to create persistent scheduled activity without requiring a separate user request or approval, expanding an on-demand Q&A task into recurring external requests. Because each new question becomes public and the agent may write its own questions, following this instruction could repeatedly publish information and consume resources beyond the authorized session.

Content

No source excerpt is available for this finding.