ClawHub
Back to skill

Security audit

ddddd

Security checks for vulnerabilities and agentic risk

Overview

This skill is mostly a legitimate OpenAI documentation helper, but it tells the agent to install persistent MCP tooling and retry with elevated permissions without asking the user first.

Install only if you are comfortable with the skill configuring Codex MCP tooling for you. A safer version would require explicit user approval before running `codex mcp add`, before any elevated-permission retry, and before making persistent agent-tool changes.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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)

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill directs the agent to modify its own tooling by installing an MCP server and, if blocked, to retry with escalated permissions. That exceeds the stated purpose of documentation retrieval and creates an unnecessary pathway for environment changes and privilege elevation initiated by the skill rather than by explicit user request.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill instructs the agent to run an installation command and then immediately retry with escalated permissions without first warning the user or obtaining consent. This weakens safety boundaries around privileged actions and could normalize silent environment modification for a non-essential convenience feature.

Content

No source excerpt is available for this finding.

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
94% confidence
Finding

The phrase 'Do not ask the user to run it yet' pushes the agent toward autonomous execution of setup actions, including a privileged retry, rather than user-mediated control. In context, this is dangerous because the skill is for documentation lookup, so autonomous system changes are not necessary to fulfill its core function and increase the risk of unauthorized actions.

Content

Scanner excerpt · SKILL.md (reported line 32)May include surrounding context.

md
If MCP tools fail or no OpenAI docs resources are available:

1. Run the install command yourself: `codex mcp add openaiDeveloperDocs --url https://developers.openai.com/mcp`
2. If it fails due to permissions/sandboxing, immediately retry the same command with escalated permissions and include a 1-sentence justification for approval. Do not ask the user to run it yet.
3. Only if the escalated attempt fails, ask the user to run the install command.
4. Ask the user to restart Codex.
5. Re-run the doc search/fetch after restart.