The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands to perform its core functions, including npm install for dependency management, and npx tsc and npx eslint for code validation. These are standard operations within a developer environment for the intended purpose of linting and type-checking.
[EXTERNAL_DOWNLOADS]: The skill downloads and installs several packages from the npm registry, such as eslint-plugin-obsidianmd, @eslint-community/eslint-plugin-eslint-comments, and various TypeScript parsers. These dependencies are required to perform the specialized Obsidian-specific code reviews described in the skill.
[INDIRECT_PROMPT_INJECTION]: The skill exposes an attack surface by ingesting and analyzing untrusted data from the user's project files.
Ingestion points: Reads content from src/*.ts, package.json, and manifest.json files within the active project directory.
Boundary markers: None; there are no explicit delimiters or instructions to ignore potential malicious prompts embedded in the source code being analyzed.
Capability inventory: The skill has the capability to write files (creating eslint.config.mjs), install software via npm, and execute arbitrary project-level code checks.
Sanitization: There is no evidence of sanitization or filtering of the source code content before it is processed by the agent or used to suggest/apply fixes.

obsidian-check