Back to skill

Security audit

Docparser

Security checks for vulnerabilities and agentic risk

Overview

This Docparser integration needs review because it treats document imports as reads and can execute downloaded installation scripts without a separate confirmation step.

Before enabling, use a reviewed CLI release and check its automatic skill-installation behavior. Require confirmation for local installation and every document import, including its URL and destination parser. Documents and account operations pass through OOMOL. These concerns support review, but 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 (6)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
94% confidence
Finding

The supplied chunk performs remote software installation rather than Docparser data operations. Installing a dependency could be a supporting implementation detail, but nothing in the supplied material establishes that this installer supports Docparser. The downloaded script is not included, so its further behavior 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 could support a Docparser integration, but the supplied chunk consists entirely of a generic login command. It neither identifies Docparser nor implements the declared data operations, so its demonstrated primary behavior differs from the stated purpose.

Content

No source excerpt is available for this finding.

External Script Fetching

High
Category
Supply Chain
Confidence
98% confidence
Finding

The command pipes a mutable network response directly into Bash without checking a pinned version, signature, or independently trusted digest; the PowerShell alternative has the same weakness. Compromise of the script delivery infrastructure could execute arbitrary code with the agent process's permissions, exposing local files and credentials; HTTPS and the legitimate dependency relationship do not eliminate that supply-chain risk.

Content

Scanner excerpt · SKILL.md (reported line 64)May include surrounding context.

The document-import action is classified as a safe read despite creating Docparser state.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The fetch_document_from_url action queues a document import and therefore changes Docparser state, but it lacks a write tag. Because the safety section explicitly treats every untagged action as a safe read, an agent could import documents without the required payload confirmation, potentially consuming quota or triggering configured downstream processing.

Content

No source excerpt is available for this finding.

The fallback setup executes downloaded scripts without warning the user about remote code execution or system changes.

Medium
Category
Output Handling
Confidence
95% confidence
Finding

A missing CLI causes the fallback instructions to direct installation through downloaded shell or PowerShell code without a separate approval step for executing code and changing the local environment. The failure-only condition limits exposure, but a Docparser request does not necessarily authorize local software installation.

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 85)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