The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill uses npx to fetch the @google/design.md package from the official NPM registry. This is a standard practice for ensuring the linter is up-to-date and targets a well-known organization's package.
[COMMAND_EXECUTION]: Common shell commands like npx, curl, and cat are used to perform validation, diffing, and token export. These commands are necessary for the skill's primary function and are used in a transparent manner.
[REMOTE_CODE_EXECUTION]: The skill executes the @google/design.md CLI tool. As this tool is provided by a reputable organization and delivered via a standard package manager, the execution is considered safe and is central to the skill's purpose.
[PROMPT_INJECTION]: The skill processes external DESIGN.md files, which presents a surface for indirect prompt injection.
Ingestion points: Content is read from local files or fetched from user-provided URLs using curl in SKILL.md.
Boundary markers: No explicit delimiters are specified for the input markdown content to isolate it from the agent's instructions.
Capability inventory: The skill possesses shell execution (npx) and network access (curl) capabilities.
Sanitization: The skill relies on the structured JSON output provided by the linter tool, which the agent then interprets for the user, providing a layer of analysis between the raw data and the final response.

design-md-validator