The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The workflow in SKILL.md instructs the agent to execute npx getdesign@latest add <slug>. This command downloads and runs code from the NPM registry at runtime, which is a high-risk pattern for remote code execution.
[EXTERNAL_DOWNLOADS]: The skill downloads the getdesign package from the public NPM registry. This dependency is not from a verified or trusted vendor, increasing the risk of supply chain attacks.
[COMMAND_EXECUTION]: The skill uses multiple shell commands, including cd, mkdir, npx, and open. The <slug> parameter in the npx command is derived from user input, which creates a potential command injection vector if the input is not strictly validated.
[PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it processes untrusted user data to generate HTML files and lacks robust sanitization.
Ingestion points: User-provided descriptions and style requests in SKILL.md.
Boundary markers: None identified; there are no instructions to ignore embedded commands within user-supplied text.
Capability inventory: Access to shell command execution (npx, open), file system writes, and external network downloads.
Sanitization: No evidence of input validation or escaping of user-provided content before it is processed or included in shell commands.

space-chart-html