obsidian-doc-workflow
Pass
Audited by Gen Agent Trust Hub on Oct 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill operates on untrusted data by reading and editing Markdown notes from user-provided paths or Obsidian vaults. This ingestion of external content creates a surface for indirect prompt injection. * Ingestion points: User-supplied .md files and vault-relative paths (SKILL.md). * Boundary markers: No explicit delimiters or 'ignore embedded instructions' markers are defined for the processed content. * Capability inventory: The skill performs file reads, surgical file edits (writes), and subprocess execution of a validation script. * Sanitization: It defines 'Obsidian Format Deltas' and 'Forbidden Cross-Platform Residue' rules to clean syntax, but lacks strict structural validation or escaping for the full document content.
- [COMMAND_EXECUTION]: The skill executes a local Python script (
.tools/skills/paper-card-delivery/scripts/validate_paper_card.py) to validate document structure. This script resides within the local project tools directory and is invoked with path arguments.
Audit Metadata