Token Usage Monitor

• Outcome: Pass
• Latest audit: May 29, 2026 · 8:21 PM UTC
• Version: 1.0.0

Overview

This is a coherent local token-usage tracker, with limited persistence and no evidence of exfiltration, credential access, or destructive behavior.

Install only if you want token usage history retained locally. Treat session IDs, model names, timestamps, and usage totals as sensitive operational metadata, and be careful before adding chat, email, or webhook reporting.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Rogue Agent: Self-Modification, Session Persistence
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (7)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 84% confidence
• Finding: The skill advertises and relies on a script that reads and writes local data (`~/.openclaw/token_usage.json`) but does not declare any permissions or boundaries for that filesystem access. Undeclared file access weakens reviewability and consent, and in an agent setting can lead to unexpected local persistence or access beyond what a user would infer from the metadata.

Missing User Warnings

Medium

• Confidence: 90% confidence
• Finding: The README encourages automatic tracking of AI model calls and storage of token usage metadata, but it does not warn users about privacy and data-governance implications. Even if only token counts are intended, the integration pattern normalizes collection around model responses and session identifiers, which can lead to retention of sensitive operational metadata without informed consent or appropriate safeguards.

Missing User Warnings

Medium

• Confidence: 88% confidence
• Finding: The README suggests extending alerts via email and webhook channels without warning that usage data may be transmitted to external systems. This can cause unintended disclosure of tenant, model, or usage information to third parties if users implement the example directly without considering authentication, redaction, and transport security.

Missing User Warnings

Medium

• Confidence: 88% confidence
• Finding: The guide recommends automatic tracking of all AI model interactions and persistent storage of usage data, but it does not mention consent, data minimization, retention, or whether prompts/metadata may reveal sensitive user activity. In this context, usage telemetry tied to session IDs and models can expose behavioral patterns and operational details, making the omission a real privacy/security concern rather than a purely documentation issue.

Missing User Warnings

Low

• Confidence: 81% confidence
• Finding: The example sends a daily usage report directly into chat without warning that token usage summaries may reveal sensitive operational patterns, model choices, or user/session activity. While the report appears limited to metrics, broadcasting it in chat can still leak information to unintended viewers or to logs retained by the platform.

Missing User Warnings

Low

• Confidence: 89% confidence
• Finding: The script persists token usage history, thresholds, session identifiers, and model activity to a local file without any notice, consent flow, or controls around sensitive metadata retention. While this is not overtly malicious, local persistence can expose behavioral data, project activity patterns, or other usage metadata to other local users, backups, or endpoint collection systems, especially because the default path is user-specific and continuously updated.

Session Persistence

Medium

• Category: Rogue Agent

Content

### scripts/

Create only the resource directories this skill actually needs. Delete this section if no resources are required.

### scripts/
- `token_usage_tracker.py`: Main script for tracking and reporting token usage

• Confidence: 88% confidence
• Finding: Create only the resource directories this skill actually needs. Delete this section if no resources are required. ### scripts/ - `token_usage_tracker.py`: Main script for tracking and reporting token

VirusTotal

66/66 vendors flagged this skill as clean.

Static analysis

No suspicious patterns detected.