Obsidian
Security checks across malware telemetry and agentic risk
Overview
The visible evidence describes an Obsidian helper that uses a third-party CLI and local Obsidian configuration in ways that appear aligned with its stated purpose, with no artifact-backed malicious behavior shown.
Before installing, verify that you trust the obsidian-cli Homebrew tap and are comfortable with the agent reading your local Obsidian configuration. I could not directly inspect the workspace artifacts because the local command sandbox failed, so this verdict is based only on the provided visible evidence and should be treated as low confidence.
VirusTotal
64/64 vendors flagged this skill as clean.
Risk analysis
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
Installing the skill may require trusting an external CLI package maintainer in addition to the skill author.
A third-party Homebrew tap introduces dependency provenance risk, but the behavior is disclosed and directly related to operating an Obsidian CLI integration.
installing a third-party command-line tool (`obsidian-cli`) from a custom Homebrew tap (`yakitrak/yakitrak`)
Review the Homebrew tap and CLI source before installing, and keep the dependency updated from a trusted source.
The agent may learn local Obsidian vault configuration details while using the skill.
The skill instructs the agent to read local Obsidian configuration. This is local file access, but it is expected for discovering Obsidian vault configuration.
read a local configuration file (`~/Library/Application Support/obsidian/obsidian.json`)
Use the skill only in workspaces where reading Obsidian configuration is acceptable, and avoid using it with vaults containing highly sensitive material unless needed.