ClawHub
Back to skill

Security audit

Clawdit

Security checks across malware telemetry and agentic risk

Overview

Clawdit is a markdown-only agent-audit skill whose file reading and proposed edits fit its stated purpose, with no hidden execution or exfiltration behavior found.

Install this only if you are comfortable letting it inspect the selected agent’s context and skill files, which may contain private preferences or operational instructions. Confirm the target workspace carefully, review any USER.md goal updates before allowing them, and inspect generated revisions or changelogs before approving changes to the target agent.

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 (3)

Vague Triggers

Medium
Confidence
93% confidence
Finding
The phrase "any variation requesting a review of their agent's configuration and belief system" leaves activation boundaries underspecified, so downstream orchestrators or agents may invoke this skill on loosely related prompts. In this skill's context, misactivation is more concerning because Phase 2 instructs directory listing and reading multiple workspace files, so an overly permissive trigger expands the chance of unnecessary sensitive-context inspection.

Vague Triggers

Medium
Confidence
93% confidence
Finding
The phrase "any variation requesting a review of their agent's configuration and belief system" leaves activation boundaries underspecified, so downstream orchestrators or agents may invoke this skill on loosely related prompts. In this skill's context, misactivation is more concerning because Phase 2 instructs directory listing and reading multiple workspace files, so an overly permissive trigger expands the chance of unnecessary sensitive-context inspection.

Vague Triggers

Low
Confidence
84% confidence
Finding
Using "or similar" in the slash-command section again broadens invocation scope without clear limits, which can cause accidental triggering from loosely related natural-language requests. This instance is somewhat less dangerous than the top-level metadata because it is supplemental, but it still reinforces ambiguous routing behavior across the skill definition.

VirusTotal

60/60 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.