The Agent Skills Directory

[PROMPT_INJECTION]: The skill features workflows that ingest untrusted external content, which constitutes a potential surface for indirect prompt injection.
Ingestion points: Data is gathered from browser-provided content (page title, URL, selection) in the web_clipper workflow and from external search results in the research_workflow.
Boundary markers: The templates defined in SKILL.md do not currently utilize explicit markers or isolation tags to separate external content from internal instructions.
Capability inventory: The skill possesses the capability to write and modify files in the local vault using the obsidian_create_note and obsidian_template tools.
Sanitization: The configuration does not specify sanitization or validation routines for data extracted from web sources before it is processed into notes.

Obsidian Automation