The Agent Skills Directory

[SAFE]: The skill serves as documentation for the Obsidian Bases file format. It does not include any executable scripts or shell commands that could compromise the host system. All documented functions and workflows are consistent with the primary purpose of data organization within a note-taking vault.
[PROMPT_INJECTION]: The skill defines a system for processing vault data that could be influenced by untrusted content in notes. Ingestion points: Data is aggregated from notes via tags, folders, and frontmatter properties. Boundary markers: The instructions do not define markers for isolating untrusted data. Capability inventory: The skill allows accessing file metadata (file()), performing logical evaluations (formulas), and custom content rendering (html()). Sanitization: The reference explicitly includes an escapeHTML() function, enabling the mitigation of cross-site scripting (XSS) or other injection risks during rendering.

obsidian-bases