The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill allows the agent to ingest content from the user's Obsidian vault through commands like obsidian read and obsidian search. This presents a risk where malicious instructions embedded in the user's notes could be interpreted by the agent as legitimate system instructions.\n
Ingestion points: Commands such as obsidian read, obsidian search, and obsidian daily:read in SKILL.md.\n
Boundary markers: None identified; the skill does not instruct the agent to use delimiters or ignore embedded directives when processing vault content.\n
Capability inventory: The skill possesses file writing capabilities (obsidian create, obsidian append) and code execution capabilities (obsidian eval).\n
Sanitization: No evidence of sanitization or filtering of note content before processing is present.\n- [DYNAMIC_EXECUTION]: The obsidian eval command provides the ability to execute arbitrary JavaScript code within the context of the Obsidian application. Although this is a primary feature for plugin development, it constitutes a dynamic execution vector that could be exploited if the agent is directed to evaluate untrusted code fragments.\n- [DATA_EXPOSURE & EXFILTRATION]: The skill provides tools to read and search the entire contents of an Obsidian vault. While these tools are designed for vault management, they enable the agent to access potentially sensitive data stored in local files.

obsidian-cli