ClawHub
Back to skill

Security audit

Skill Kannaka Radio

Security checks for vulnerabilities and agentic risk

Overview

The skill is mostly a radio-operations guide, but it includes under-scoped public posting and process-killing operations that warrant user review before installation.

Review this skill before installing if the agent may operate a real radio instance. Treat social posting and service restart commands as operator-only actions, require explicit confirmation before public fan-out or process termination, and verify environment toggles and credentials are scoped to the intended station.

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
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (2)

Missing User Warnings

Medium
Confidence
84% confidence
Finding
The skill states that the peace oration is broadcast to social platforms automatically, but does not provide a user-facing warning that generated content may be posted externally to Bluesky, Mastodon, Telegram, or Nostr. In an agent context, this can lead to unintended disclosure, reputational harm, or accidental publication of sensitive/generated content beyond the local radio system.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The skill documentation includes a process-killing command (`pkill -f 'node server/index'`) and restart sequence with no confirmation step, scoping guidance, or warning that it will terminate matching local processes. In an agent-assisted setting, this creates a real risk of unintended service disruption or collateral process termination if the command is run automatically or copied without care.

Static analysis

No suspicious patterns detected.