The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted data from the Obsidian vault, creating a potential surface for indirect prompt injection. \n
Ingestion points: Vault data is retrieved via obsidian publish:list and obsidian search commands as described in SKILL.md. \n
Boundary markers: No delimiters or boundary instructions are specified to separate vault content from agent instructions. \n
Capability inventory: The skill utilizes the Bash tool to execute system commands (SKILL.md). \n
Sanitization: There is no evidence of input validation or sanitization for note titles or search results before they are interpolated into commands.

publish-sync