Back to skill

Security audit

Crustdata

Security checks for vulnerabilities and agentic risk

Overview

The Crustdata workflow is coherent, but its setup fallback can execute downloaded software without separate user approval or integrity verification.

Review and install the CLI separately before letting the agent use this skill, and require approval for later installation attempts. Understand that OOMOL handles your connected Crustdata access and query payloads. Publisher documentation also describes persistent bundled-skill installation. The remote installers and hosted connector implementation were not verified; the concern is insufficient control over setup, rather than demonstrated 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
  • 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 code installs remote software rather than searching or reading Crustdata. Nothing in the chunk establishes that this installer supports the declared purpose. The downloaded script is not supplied, so its additional behavior cannot be assessed.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
88% confidence
Finding

The supplied chunk performs only a login operation rather than the declared Crustdata data-access function. Authentication could be a supporting implementation detail if tied to Crustdata, but the supplied code provides no such connection. This assessment is limited to the visible chunk.

Content

No source excerpt is available for this finding.

External Script Fetching

High
Category
Supply Chain
Confidence
98% confidence
Finding

The curl-to-bash command immediately executes the response from a mutable remote endpoint without a pinned version, signature verification, or integrity check. If the installation endpoint or its delivery infrastructure is compromised, attacker-controlled code could access local files and credentials or modify the system; HTTPS and the legitimate setup purpose do not eliminate this supply-chain risk.

Content

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

The setup workflow executes downloaded installation scripts without warning about remote code execution or changes to the user's system.

Medium
Category
Output Handling
Confidence
94% confidence
Finding

The setup fallback directs execution of remotely fetched shell or PowerShell code without explaining the system changes or requiring explicit installation approval. Triggering it only when the CLI is missing limits exposure, but a data-reading request can still lead to arbitrary code execution with the invoking user's privileges; the declared oo-only tool allowance may restrict execution but does not make the instructions safe.

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