Back to skill

Security audit

Obsidian

Security checks for vulnerabilities and agentic risk

Overview

The verdict is benign because the scanners found no reported malicious behavior, though they flagged capabilities beyond the stated Obsidian vault-management purpose and some artifact references could not be resolved.

Before installing, decide whether you need note deletion, moving, and screenshot capture, and whether you are comfortable granting access for those actions. The scan was partial because it could not resolve several references, so the available evidence is incomplete.

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
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
92% confidence
Finding

The commands use the declared CLI and operate on the expected vault resources. However, the supplied chunk performs relocation and deletion rather than the listed read/search/create/edit operations. Deletion is a material undeclared capability, not merely a supporting implementation detail.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
88% confidence
Finding

Plugin reloading is explicitly within scope, and error inspection can support plugin work. The evaluated JavaScript only counts vault files, which is consistent with vault inspection. However, capturing the application UI accesses and exports visual content beyond the declared note, task, link, property, and plugin operations.

Content

No source excerpt is available for this finding.