obsidian-structure

Pass

Audited by Gen Agent Trust Hub on Mar 12, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection as it processes untrusted data from the user's Obsidian vault. \n
  • Ingestion points: The scripts/structure.py script recursively reads every markdown file (.md) in the vault to parse frontmatter and wikilinks. \n
  • Boundary markers: The script output is provided to the AI context without protective delimiters or instructions to ignore embedded commands found within the processed notes. \n
  • Capability inventory: The AI agent is authorized to create and manage files across multiple vault directories, creating a significant action surface if an injection occurs. \n
  • Sanitization: No sanitization is performed on the filenames, tags, or links extracted from the notes before they are presented to the AI context.\n- [COMMAND_EXECUTION]: The skill executes a local Python script to analyze the vault structure. \n
  • Evidence: The SKILL.md file defines a command to run python3 .claude/skills/obsidian-structure/scripts/structure.py. While the path is local and the script is bundled with the skill, it allows the agent to execute code that accesses the local filesystem.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 12, 2026, 11:42 PM