obsidian-vault-health-sweep
Pass
Audited by Gen Agent Trust Hub on May 19, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to Indirect Prompt Injection because it ingests untrusted data from an Obsidian vault and has the capability to modify the file system.
- Ingestion points: Reads local Markdown notes to identify broken links, tags, and frontmatter drift across multiple stages.
- Boundary markers: The instructions do not define boundary markers or explicit prompts to ignore instructions found within the vault content.
- Capability inventory: Includes file-write operations (updating frontmatter, links, and tags), file-creation (creating stub notes), and potentially file-deletion (removing deleted note references).
- Sanitization: There is no mention of sanitizing or validating the content of the notes before the agent processes them, which could allow malicious instructions embedded in a note to influence the agent's behavior during the 'Health Sweep'.
Audit Metadata