ClawHub
Back to skill

Security audit

Skill Tracker

Security checks for vulnerabilities and agentic risk

Overview

This skill is a local OpenClaw skill-usage tracker whose file access and persistent logs match its stated purpose, but users should avoid storing sensitive metadata.

Install only if you want local telemetry about skill usage. Keep metadata minimal, do not log prompts, credentials, customer data, tokens, or sensitive file contents, and add the optional cron jobs only if you want scheduled ongoing analysis.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • 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 (4)

Missing User Warnings

Medium
Confidence
88% confidence
Finding
The README encourages automatic logging of skill usage, timing, satisfaction, and arbitrary metadata, but does not describe data minimization, consent, retention, or what sensitive values may end up in metadata. In a multi-skill agent environment, this can lead to unintentional collection of prompts, identifiers, or operational telemetry that users and operators did not expect to persist.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill advertises automatic tracking of all skill usage, success/failure, timing, and satisfaction without any privacy notice, consent flow, retention limit, or data minimization guidance. In practice, operational telemetry and satisfaction fields can reveal user behavior patterns and sensitive workflow details, especially when collected across all skills.

Missing User Warnings

Medium
Confidence
97% confidence
Finding
The example encourages callers to log arbitrary metadata via a free-form metadata object, with no warning against including prompts, tokens, personal data, secrets, or file contents. That pattern can easily lead integrators to persist highly sensitive data into local logs, creating a durable disclosure risk and expanding the blast radius of any later compromise or accidental sharing.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The storage section documents persistent log, score, and proposal directories under a user home path but provides no warning about privacy implications, access control, encryption, retention, or cleanup. Persistent analytics artifacts can accumulate sensitive operational history over time, making them valuable to insiders, malware, or anyone with local access.

Static analysis

No suspicious patterns detected.