skills/beita6969/scienceclaw/obsidian/Gen Agent Trust Hub

obsidian

Pass

Audited by Gen Agent Trust Hub on Apr 6, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill installs the third-party utility obsidian-cli via the Homebrew package manager from the yakitrak/yakitrak repository.
  • [COMMAND_EXECUTION]: Utilizes shell commands and the obsidian-cli binary to perform file operations such as searching, creating, renaming, and deleting Markdown notes within the local file system.
  • [COMMAND_EXECUTION]: Reads the local application configuration file located at ~/Library/Application Support/obsidian/obsidian.json to resolve vault paths and identify the active workspace. This is a standard functional requirement for interacting with the Obsidian desktop application.
  • [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it ingests untrusted data from Markdown files.
  • Ingestion points: Reads *.md (plain text Markdown) files from the user's local Obsidian vault.
  • Boundary markers: No explicit delimiters or boundary instructions are used when processing note content.
  • Capability inventory: Includes file system writes, deletions, and search operations via obsidian-cli and direct file edits.
  • Sanitization: No evidence of content sanitization or validation before processing note data.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 6, 2026, 11:30 PM
Security Audit — agent-trust-hub — obsidian