ClawHub
Back to skill

Security audit

Discord Cross-Gateway Delegation

Security checks for vulnerabilities and agentic risk

Overview

This skill is a documented Discord bot delegation guide, but it forwards DM content and identifiers to another bot/channel with weak consent and scoping guidance.

Review before installing. Use this only for trusted operator-controlled Discord bots and private worker lanes. Make sure users know when their DM content and Discord identifiers will be forwarded to another bot, restrict who can access the worker lane, and require confirmation for browser, credential, ops, or external-action tasks.

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 (3)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The trigger list is unusually broad and includes multiple generic natural-language phrases plus opaque aliases like 'maekjini' and 'kaijini', which can cause the skill to activate in unintended contexts. In a cross-gateway delegation skill, accidental activation is more dangerous than normal because it may initiate message relays or task delegation to an external bot, increasing the chance of unintended data sharing or unauthorized workflow execution.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The process explicitly instructs the controller bot to copy DM-derived metadata (source-channel-id, source-message-id, requester-discord-id) and the worker's final answer into a separate worker lane and then relay content back into the original DM. That is cross-channel movement of potentially sensitive user data, and the document does not require explicit user notice, consent, minimization, or access restrictions beyond soft recommendations like 'owner/operator scope only when possible.'

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The quick handoff mode allows delegation based on a simple worker-name prefix and explicitly prioritizes speed, but it does not define any scope limits, authorization checks, or confirmation requirements. In a cross-gateway Discord delegation setup, that ambiguity can cause unintended task forwarding, prompt injection via natural-language phrasing, or execution of higher-risk actions on the remote worker without sufficient controller validation.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.