The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill exposes the obsidian eval command, which allows the execution of arbitrary JavaScript code directly within the Obsidian application's Electron context. This capability can be leveraged to perform actions outside the scope of note management, such as accessing the file system or network with the application's privileges.
[DYNAMIC_EXECUTION]: The obsidian dev:cdp command provides access to the Chrome DevTools Protocol, allowing low-level manipulation and debugging of the application instance.
[EXTERNAL_DOWNLOADS]: The skill includes commands to install community plugins (obsidian plugin:install) and themes (obsidian theme:install). These commands download and integrate third-party code from external sources into the local Obsidian environment.
[INDIRECT_PROMPT_INJECTION]: The skill contains several commands that read data from the Obsidian vault, creating an attack surface for indirect prompt injection if the vault content is untrusted.
Ingestion points: Commands like obsidian read, obsidian search, obsidian daily:read, and obsidian backlinks ingest user-controlled file content into the agent's context (found in SKILL.md).
Boundary markers: None provided in the command instructions or skill definition.
Capability inventory: The skill possesses significant write and execution capabilities, including obsidian create, obsidian eval, obsidian plugin:install, and obsidian delete (found in SKILL.md).
Sanitization: There is no evidence of sanitization or filtering of the content retrieved from the vault before it is processed by the agent.
[COMMAND_EXECUTION]: The obsidian command tool allows the agent to trigger any internal Obsidian command by ID, providing broad access to application functionality.

obsidian