Back to skill

Security audit

SheetDB

Security checks for vulnerabilities and agentic risk

Overview

The SheetDB features are coherent, but automatic setup can execute unverified remote code and introduce broader changes to your AI agents.

Review the setup behavior before installing. Install a verified CLI release yourself and decide whether it may synchronize additional agent skills. The [command reference](https://github.com/oomol-lab/oo-cli/blob/main/docs/commands.md) documents OO_SKILLS_SYNC_DISABLED=1 to disable startup synchronization. Connect only the intended SheetDB account and retain approval for data changes. The remote installer contents remain unverified; no malicious behavior 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 (5)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The supplied command executes an Oomol-hosted installer rather than performing SheetDB operations. Installing unrelated software is not declared, and the chunk provides no evidence that it is a supporting dependency for SheetDB functionality. The downloaded script's contents and subsequent effects cannot be determined from this chunk.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding

The supplied code only invokes a CLI login command and shows no SheetDB operations or integration. Authentication could support a SheetDB implementation, but nothing in this chunk establishes that relationship, so its observable purpose does not match the declared SheetDB functionality.

Content

No source excerpt is available for this finding.

External Script Fetching

High
Category
Supply Chain
Confidence
98% confidence
Finding

The command streams a mutable remote script directly into Bash without a pinned version or independently verified integrity; the PowerShell alternative uses the same download-and-execute pattern. Compromise of the distribution endpoint could execute attacker-controlled code with the invoking user's privileges, potentially exposing local files and credentials; HTTPS alone does not protect against a compromised source.

Content

Scanner excerpt · SKILL.md (reported line 62)May include surrounding context.

First-time setup instructs the agent to execute downloaded installation scripts without warning the user about local system changes.

Medium
Category
Output Handling
Confidence
94% confidence
Finding

After a missing-command error, the skill instructs installation through remotely downloaded scripts without requiring approval for executing code or changing the local environment. The conditional setup reduces exposure, but a legitimate SheetDB request could still trigger installation outside the user's intended data-operation scope if the agent's tool permissions permit it.

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 83)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