ClawHub
Back to skill

Security audit

LLM Monitor

Security checks across malware telemetry and agentic risk

Overview

This skill coherently helps check LLM usage through clawmeter, with some sensitive setup and daemon actions that are disclosed and mostly gated by user consent.

Install only if you want an agent to run local clawmeter commands for usage checks. Review prompts carefully before approving package installation, daemon start, systemd service installation, or history purge. Do not paste API keys into chat; configure credentials directly in your local clawmeter setup.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Memory PoisoningPersistent Context Injection, Context Window Stuffing, Memory Manipulation
  • Rogue AgentSelf-Modification, Session Persistence
  • 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
Findings (4)

Vague Triggers

Medium
Confidence
93% confidence
Finding
The trigger conditions are very broad and match common, everyday phrases such as asking about usage, costs, or limits. In an agent environment, this can cause the skill to activate unexpectedly and perform local system checks, configuration changes, or daemon actions when the user did not clearly intend to invoke this specific capability.

Memory Manipulation

High
Category
Memory Poisoning
Content
| View daemon logs | Check `~/.local/state/clawmeter/daemon.log` |
| Database stats | `clawmeter history stats` |
| Export history | `clawmeter history export --format csv` |
| Purge history | `clawmeter history purge` (ask user first — destructive) |

---
Confidence
90% confidence
Finding
The skill includes a destructive history purge capability, which can delete collected usage data. Although it says to ask the user first, broad triggering or ambiguous user prompts could still lead to irreversible data loss, especially if the agent over-interprets maintenance requests.

Session Persistence

Medium
Category
Rogue Agent
Content
Would you like me to start the daemon for you?
```

If the user agrees, run `clawmeter daemon start`. For systemd installation, ask for confirmation first since it modifies system services.

### 1d. Quick health check
Confidence
89% confidence
Finding
The skill allows starting a daemon and installing it as a system service, which introduces persistence on the host. Even with a confirmation requirement for systemd installation, this meaningfully changes system state and can outlive the session, making unintended invocation more dangerous.

Session Persistence

Medium
Category
Rogue Agent
Content
test -f "${CLAWMETER_CONFIG:-$HOME/.config/clawmeter/config.toml}" && echo "exists" || echo "missing"
```

If missing, create a minimal skeleton config (with no secrets):

```bash
mkdir -p ~/.config/clawmeter
Confidence
84% confidence
Finding
Automatically creating configuration directories and files writes persistent state to the user's home directory. While the config is described as containing no secrets, this still modifies the environment and may prime later automated actions, especially when combined with broad triggers.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.