obsidian-markdown
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill consists entirely of documentation and formatting examples. It does not contain any executable scripts, binaries, shell commands, or network requests.
- [INDIRECT_PROMPT_INJECTION]: As a document-processing skill intended for use with Obsidian vaults, the agent may encounter untrusted content within notes (such as markdown comments or callouts) that could contain malicious instructions.
- Ingestion points: Obsidian markdown (.md) files as specified in the skill's name and description.
- Boundary markers: Absent; the skill defines formatting syntax but does not provide instructions for delimiting or ignoring untrusted instructions embedded in files.
- Capability inventory: The skill itself defines no tools, but is designed for use by an agent with file-system read/write capabilities.
- Sanitization: None present; the skill focused on syntax definition rather than input validation or sanitization.
Audit Metadata