Back to skill

Security audit

VectorShift

Security checks for vulnerabilities and agentic risk

Overview

The VectorShift workflow is coherent, but its setup fallback directs the agent to execute unverified downloaded code without a separate approval step.

Review the setup behavior before installing. Install a reviewed, version-fixed CLI separately and require approval before any downloaded installer runs. Review the CLI's automatic agent-skill synchronization and the VectorShift permissions entrusted to OOMOL. Installer contents and the hosted connector implementation were unavailable for verification; the inspected artifact does not establish malicious intent.

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 (5)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The supplied command performs remote script execution for an apparent Oomol CLI installation and contains no VectorShift operations. No relationship is shown that would establish this installation as a supporting implementation detail for VectorShift. The downloaded script's contents and further effects cannot be determined from the supplied code.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
87% confidence
Finding

The supplied code consists solely of a CLI login command rather than a VectorShift integration. Authentication could be a supporting implementation detail if the CLI were established as a VectorShift tool, but the supplied content provides no such connection. As shown, its primary behavior does not substantiate the declared purpose.

Content

No source excerpt is available for this finding.

External Script Fetching

High
Category
Supply Chain
Confidence
99% confidence
Finding

Piping an HTTPS download directly into bash executes mutable remote content without inspecting it or checking a pinned digest or signature. HTTPS protects transport but does not protect against compromise of the hosting service or publisher, which could turn this legitimate dependency-installation step into arbitrary code execution under the invoking account.

Content

Scanner excerpt · SKILL.md (reported line 58)May include surrounding context.

Fallback setup instructs the agent to execute a downloaded installer without warning the user or obtaining approval.

Medium
Category
Output Handling
Confidence
97% confidence
Finding

The missing-CLI fallback instructs installation through immediate execution of downloaded shell or PowerShell code without requiring user approval or integrity verification. Although limited to first-time setup, it expands an ordinary VectorShift request into local code execution and system modification; compromised installer delivery could execute attacker-controlled code with the agent's privileges.

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 79)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