The Agent Skills Directory

[COMMAND_EXECUTION]: The skill exposes an eval command (obsidian eval code="...") which facilitates the execution of arbitrary JavaScript within the context of the running Obsidian application.
[COMMAND_EXECUTION]: The dev:cdp command enables interaction with the Chrome DevTools Protocol, allowing for low-level application manipulation and data access.
[EXTERNAL_DOWNLOADS]: The plugin:install command allows the agent to download and install external code (Obsidian plugins) from the community marketplace at runtime.
[EXTERNAL_DOWNLOADS]: The documentation includes instructions for installing the obsidian-headless package globally via npm, which introduces an external dependency.
[PROMPT_INJECTION]: The skill is susceptible to Indirect Prompt Injection (Category 8) due to its data-processing capabilities.
Ingestion points: The agent can ingest untrusted data from the vault using obsidian read, obsidian template:read, and obsidian search.
Boundary markers: There are no documented delimiters or instructions to treat file content as untrusted data.
Capability inventory: The skill provides high-impact capabilities including obsidian eval, obsidian dev:cdp, obsidian plugin:install, and the ability to delete or move files.
Sanitization: No sanitization or validation of data retrieved from the vault is mentioned before being used in potentially dangerous operations.

obsidian-cli