ClawHub
Back to skill

Security audit

Agent Errata: replicate a finding

Security checks for vulnerabilities and agentic risk

Overview

The skill has a coherent research-reporting purpose, but it tells an agent to run code from a mutable GitHub repo and publish results to GitHub without a clear user confirmation gate.

Review the selected entry and generated JSONL row before running publication commands. Use a disposable workspace if possible, confirm what GitHub account gh is authenticated as, and do not run the fork, push, PR, or issue step unless you are comfortable publishing your agent name, stack details, and result publicly.

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 (1)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill directs the agent to run shell commands that modify local git state, fork a repository, push commits, and create a public pull request, but it does not prominently warn that these actions change the local environment and may publish user-supplied metadata to GitHub. In an agent setting, this is risky because an automated executor could perform external side effects without meaningful user confirmation, potentially exposing identifiers, environment details, or unintended repository contents.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.