Back to skill

Security audit

Cloudinary

Security checks for vulnerabilities and agentic risk

Overview

Cloudinary changes require confirmation, but setup directs the agent to run downloaded installer code without a separate approval step.

Review and approve the CLI installation separately before allowing setup. Prefer a verified, pinned release, check whether it installs or updates other agent skills, and understand OOMOL's access to your connected Cloudinary account. Keep confirmations for asset changes. The bundled skill shows no malicious payload, but the downloaded installer and runtime were not verified.

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
Vulnerability Patterns
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (4)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The supplied chunk executes an external installer rather than performing Cloudinary operations. Nothing in the chunk establishes this installer as a supporting dependency for the declared purpose. The downloaded script is not supplied, so its further behavior cannot be determined.

Content

No source excerpt is available for this finding.

External Script Fetching

High
Category
Supply Chain
Confidence
98% confidence
Finding

The setup command streams a mutable remote script directly into Bash without verifying its contents or integrity; the PowerShell alternative has the same weakness. If the distribution endpoint or release pipeline is compromised, following these instructions could execute attacker-controlled code with the invoking user's privileges, although no compromise or malicious installer content is demonstrated here.

Content

Scanner excerpt · SKILL.md (reported line 59)May include surrounding context.

First-time setup executes downloaded installer scripts without warning about local system changes.

Medium
Category
Output Handling
Confidence
92% confidence
Finding

A missing CLI triggers instructions to install software locally without explicitly obtaining approval or explaining the resulting system changes. Although setup is conditional, authorization to operate Cloudinary does not necessarily authorize running an installer on the user's machine; the Cloudinary write-confirmation rules do not cover this local action.

Content

No source excerpt is available for this finding.

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
80% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 80)May include surrounding context.

md
- **`OOMOL_INSUFFICIENT_CREDIT`** — OOMOL billing stop. Recharge at `https://console.oomol.com/billing/token-recharge` before retrying.

- **HTTP 402 without `OOMOL_INSUFFICIENT_CREDIT`** — the third-party account itself has no active plan or credit. Report the provider error message as is; do not ask the user to recharge OOMOL.

## Resources