ClawHub
Back to skill

Security audit

Skill Agent Broadcast

Security checks for vulnerabilities and agentic risk

Overview

This skill is a real broadcast tool, but it can send messages to multiple Telegram or Discord group targets by default and should be reviewed before installation.

Install only if you intend an agent to send messages to these groups. Before use, replace config/groups.json with your own approved targets, avoid relying on the default all-groups behavior, and provide a least-privilege OPENCLAW_TOKEN for the local gateway.

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
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • 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)

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding
The supplied code is only a message-formatting utility. It constructs formatted strings for several alert types and exports them. There is no logic for sending messages, enumerating group sessions, coordinating agents, or interacting with any OpenClaw messaging APIs. Because the declared purpose centers on broadcasting to multiple sessions, while the actual code only prepares message content, the description materially overstates the implemented behavior.

Lp3

Medium
Category
MCP Least Privilege
Confidence
90% confidence
Finding
The skill declares executable capabilities and dependence on environment variables, including an auth token, but does not define any explicit tool scope or permission boundaries. In a broadcast skill that can send messages to multiple external group sessions, this omission increases the chance of unauthorized use, secret exposure, or unintended mass messaging because agents may invoke it without clear restrictions.

Static analysis

Detected: suspicious.env_credential_access

Environment variable access combined with network send.

Critical
Code
suspicious.env_credential_access
Location
scripts/broadcast.js:25