ClawHub
Back to skill

Security audit

Obsidian Vault Linker

Security checks for vulnerabilities and agentic risk

Overview

This skill transparently helps an agent read and update Obsidian notes with typed links, with approval required by default before writing.

Install this only for vaults you are comfortable letting an agent read. Keep the default review-before-write workflow for important notes, and use autonomous mode only on a scoped folder or after backing up/versioning the vault.

A.I.G

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

SkillSpector

Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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)

Context-Inappropriate Capability

Medium
Confidence
92% confidence
Finding
The skill explicitly instructs the agent to run local shell commands (`which obsidian`, `obsidian version`, and other `obsidian ...` commands) during normal operation. That expands the skill from simple file reading/writing into command execution against the host environment, which increases attack surface and can violate least-privilege expectations for a vault-linking skill.

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
2. Link spoke notes into hubs before looking for lateral spoke-to-spoke connections
3. Hub-to-hub relationships (e.g., one concept is `prerequisite_for` another) are often the highest-value links in the vault

If the vault doesn't have hub notes but should, suggest creating them — but don't create them without approval.

### Step 2: Read and Summarize
Confidence
75% 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.

Static analysis

No suspicious patterns detected.