Back to skill

Security audit

RenderForm

Security checks for vulnerabilities and agentic risk

Overview

This skill needs review because it treats potentially billable output creation as safe reading and directs automatic installation of unverified remote code.

Before installing, require confirmation for rendering and screenshots, including their payloads and possible credit costs. Review and verify the CLI installation separately. Its documentation describes automatic changes to installed agent skills and provides OO_SKILLS_SYNC_DISABLED to disable startup synchronization. [CLI controls](https://github.com/oomol-lab/oo-cli/blob/main/docs/commands.md) The remote installer and hosted connector implementation were not verified; no bundled evidence establishes malicious behavior.

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
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (6)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The supplied command executes an external installation script, with no visible RenderForm interaction or demonstrated connection to the declared task. This is a materially different capability, not an established supporting implementation detail. The downloaded script is not supplied, so its additional effects cannot be assessed.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
86% confidence
Finding

Authentication can be a supporting implementation detail, but the supplied chunk consists entirely of a generic CLI login command and does not establish that it authenticates to RenderForm or supports the declared data-access workflow. Its visible primary behavior therefore differs from the declared purpose.

Content

No source excerpt is available for this finding.

External Script Fetching

High
Category
Supply Chain
Confidence
97% confidence
Finding

The installation command streams a mutable remote script directly into Bash without pinning or independent integrity verification; the PowerShell alternative has the same issue. Compromise of the distribution endpoint could execute attacker-controlled code with the agent process's privileges, although the missing-CLI prerequisite and any enforced tool restrictions limit exposure.

Content

Scanner excerpt · SKILL.md (reported line 60)May include surrounding context.

The skill exposes content-generation actions despite describing its scope as searching and reading data.

Medium
Category
Excessive Agency
Confidence
96% confidence
Finding

The declared search-and-read scope includes image rendering and screenshot creation, which create generated outputs rather than merely retrieve existing data. In combination with the untagged-action safety rule, this scope expansion can authorize generation and associated credit consumption without the safeguards users would expect from a read-oriented skill.

Content

No source excerpt is available for this finding.

Generation actions are classified as safe reads, bypassing the documented confirmation requirement.

Medium
Category
Excessive Agency
Confidence
98% confidence
Finding

Neither render_image nor take_screenshot carries a write tag, while the safety section explicitly declares all untagged actions safe reads. This bypasses the skill's confirmation requirement for operations that create outputs and may consume account credits, making the missing tags operationally significant.

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 81)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