obsidian

Pass

Audited by Gen Agent Trust Hub on Mar 30, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill installs the obsidian-cli tool using Homebrew from a third-party repository (yakitrak/yakitrak/obsidian-cli).- [COMMAND_EXECUTION]: The agent executes shell commands through obsidian-cli to perform file operations, including searching, creating, moving, and deleting Markdown notes within the vault.- [PROMPT_INJECTION]: Processing user-controlled Markdown notes creates a surface for indirect prompt injection where malicious text in a note could attempt to override agent behavior.
  • Ingestion points: Note content retrieved via search-content.
  • Boundary markers: No markers are used to separate note data from instructions.
  • Capability inventory: Tooling allows for file deletion and modification.
  • Sanitization: No content filtering or validation is performed on the notes read.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 30, 2026, 12:15 AM