ClawHub
Back to skill

Security audit

validation-agent

Security checks for vulnerabilities and agentic risk

Overview

This skill is a straightforward validation checklist for reviewing code changes, with expected but non-trivial use of local test and audit commands.

Install only if you want an agent to perform code validation tasks. Before allowing it to run tests, builds, benchmarks, or dependency audits on an unfamiliar repository, review the project scripts and prefer a sandbox or disposable environment.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

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 (3)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill directs the agent to run project-local test commands like npm test, pytest, and similar via Bash without any warning that these can execute arbitrary repository-defined scripts, hooks, or setup code. In the context of an agent skill, this is dangerous because a malicious or compromised repository can abuse normal validation steps to run attacker-controlled code on the user's machine or CI environment.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill recommends audit commands such as npm audit, pip-audit, and cargo audit with no warning that they may invoke external binaries, access the network, download advisory data, or modify caches and the execution environment. For an automated agent, this omission can lead to unintended outbound connections or execution of tooling in sensitive environments where network use or environmental changes must be controlled.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The command reference section broadly enumerates executable commands across multiple ecosystems without a general warning about side effects, despite many of them running repository-controlled scripts or external tooling. This increases risk because users or downstream agents may treat the section as safe boilerplate and execute commands that build, lint, test, or audit untrusted projects, enabling code execution or environmental impact.

Content

No source excerpt is available for this finding.