ClawHub
Back to skill

Security audit

Publish Audit

Security checks for vulnerabilities and agentic risk

Overview

This is a documentation-only skill for auditing ClawHub skill folders before publishing, with no executable code, network use, credentials, or persistence.

Reasonable to install for pre-publish checks. Be aware it may give audit-style advice about local skill contents, including possible secret references, so use it on folders you intend to review and do not treat it as a full malware or install-safety vetter.

SkillSpector

Vulnerability Patterns
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (1)

Vague Triggers

Medium
Confidence
95% confidence
Finding
The frontmatter description is topic-oriented and does not define clear invocation boundaries or 'do not use for' conditions, even though the skill itself emphasizes that such boundaries are necessary. This can cause the agent to invoke the auditing skill in inappropriate contexts, including reviewing untrusted skills or general security tasks it was not designed for, which increases the chance of misleading results and unsafe automation.

Static analysis

No suspicious patterns detected.