Back to skill

Security audit

Leina

Security checks for vulnerabilities and agentic risk

Overview

The Leina workflow has sensible safeguards, but its automatic setup runs unverified downloaded code without a clear approval step.

Review the CLI installation separately and approve a verified version before allowing setup. Review or disable automatic installation of additional agent skills. Confirm the robot, recipients, and message before sending notifications. The complete bundled skill was inspected, but the remote installers and connector implementation remain unverified; no malicious intent was established.

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
97% confidence
Finding

The supplied code installs software from Oomol; it does not show any Leina integration or data operations. Although installation can be a supporting implementation detail, no connection between this installer and the declared Leina functionality is established. The downloaded script's contents are unavailable, so its further effects cannot be determined from this chunk.

Content

No source excerpt is available for this finding.

External Script Fetching

High
Category
Supply Chain
Confidence
98% confidence
Finding

Piping the HTTPS response directly to bash executes mutable remote code with the agent process's privileges before it can be inspected or integrity-checked. The documented OOMOL dependency makes the destination relevant, not inherently malicious, but compromise of that distribution path could expose local files and credentials or alter the host; this overlaps with SQP-2.

Content

Scanner excerpt · SKILL.md (reported line 56)May include surrounding context.

Fallback setup instructs the agent to execute remote installation scripts without a system-integrity warning.

Medium
Category
Output Handling
Confidence
95% confidence
Finding

The fallback directs the agent to execute remotely fetched shell or PowerShell code without first reviewing or verifying it or obtaining explicit approval for installation. Although conditional on a missing CLI and relevant to setup, this exposes the host to arbitrary code execution if the installer source or delivery infrastructure is compromised.

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 71)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