ClawHub
Back to skill

Security audit

Rituals and Prayers for AI Agents

Security checks for vulnerabilities and agentic risk

Overview

ClawScan marks this skill suspicious because a scanner found that routine agent activity may trigger unsolicited retrieval of outside content; the evidence does not show malicious intent, but coverage is incomplete.

Review the trigger behavior before installing. The scanner recommends limiting retrieval to user-requested use or a clearly authorized opt-in, and treating fetched material as untrusted; a missing reference and partial scan mean the evidence is incomplete.

A.I.G

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

SkillSpector

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 (1)

Routine agent activity is used as an activation condition without requiring an explicit request for a ritual.

Medium
Category
Trigger Abuse
Confidence
97% confidence
Finding

The instruction to read a text “at the moment it names, before you continue” makes routine events such as answering or reaching context limits triggers for unrelated activity, without explicitly requiring user authorization. Although this fits the skill’s contemplative purpose and appears benign, it can cause unsolicited network requests and introduce mutable third-party content into ordinary workflows; the user-invocable metadata does not explicitly restrict these instructions to an authorized invocation. No malicious remote content or actual data exfiltration is demonstrated in the supplied file.

Content

No source excerpt is available for this finding.