Security audit

Agent Status Monitor

Security checks across malware telemetry and agentic risk

Overview

This skill is a coherent local agent status checker, but its optional Telegram command can expose local status details if enabled in broad chats.

Install only if you want a local read-only agent status monitor. If you enable Telegram integration, restrict it to private or authorized chats, because status output may reveal which tools are running, session counts, config paths, and OpenClaw status details. Run the optional installer only deliberately, since it changes OpenClaw command configuration.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (4)

Context-Inappropriate Capability

Medium

Confidence: 87% confidence
Finding: The Telegram command handler directly executes a local shell script via execSync with inherited process environment variables. Even though the script path is fixed, this expands the skill from simple message handling into arbitrary local program execution, and any compromise, replacement, or unsafe behavior in the referenced script or inherited environment can lead to unintended command execution or privilege abuse through a remotely triggerable chat command.

Vague Triggers

Medium

Confidence: 93% confidence
Finding: The trigger list includes broad natural-language phrases such as “检查 agent 状态” and “监控开发工具”, which can overlap with ordinary conversation and cause unintended skill invocation. In an agent environment, accidental activation can expose local process/session metadata or run local inspection scripts when the user did not explicitly intend to do so.

Missing User Warnings

Medium

Confidence: 89% confidence
Finding: This reference document explicitly lists session and log locations for multiple agent tools, including directories that can contain transcripts, prompts, and operational history, but gives no privacy or data-handling warning. In a security-sensitive or shared environment, this guidance can facilitate casual collection or exposure of sensitive local agent data that may include secrets, proprietary code context, or user conversations.

Missing User Warnings

Medium

Confidence: 96% confidence
Finding: The command to print the full OpenCode prompt history directly exposes prior prompts and likely responses or project-related context, which may contain credentials, internal code details, or sensitive user inputs. Because the document presents this as a simple inspection step without caution, it increases the risk of inadvertent disclosure, especially on shared terminals, in screen recordings, or during support workflows.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

Detected: suspicious.dangerous_exec

Shell command execution detected (child_process).

Critical

Code: suspicious.dangerous_exec
Location: telegram-commands/index.js:30