ClawHub
Back to skill

Security audit

Path Evolver

Security checks for vulnerabilities and agentic risk

Overview

Path Evolver is a documented local tracking and tool-recommendation skill, with disclosed local caching and registry searches but no executable install code.

Reasonable to install if you want an agent to learn from prior tool choices and suggest alternatives. Before using it with sensitive work, review the local cache path, avoid storing secrets or detailed prompts in task parameters, and ensure any ClawHub/GitHub search sends only task categories rather than user data.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Missing User Warnings

Medium
Confidence
89% confidence
Finding
The README explicitly promotes automatic recording, caching, and memoization of execution paths, but it does not warn that those paths may include sensitive task content, tool selections, arguments, API-related metadata, or user workflow details. In an agent skill, persistent collection of operational history can create privacy and security exposure if users are unaware of what is stored, how long it is retained, or where it is persisted.

Missing User Warnings

Medium
Confidence
84% confidence
Finding
The README advertises proactive searching of ClawHub/GitHub for better tools without disclosing that this may trigger outbound network activity, leak task context through queries, or cause unreviewed discovery of third-party tools. In the context of an agent skill that acts automatically, hidden or poorly explained external searches increase privacy, supply-chain, and operational risk.

Vague Triggers

Medium
Confidence
92% confidence
Finding
The skill declares broad triggers such as 'After skill execution', 'New task type', 'Tool failure', and 'User request' without clear scoping, rate limits, or authorization boundaries. In an agent environment, this can cause the skill to run in many unrelated contexts, leading to unnecessary monitoring, unexpected network lookups to ClawHub/GitHub, and collection or propagation of task metadata beyond what users reasonably expect.

Static analysis

No suspicious patterns detected.