The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill is a "catalogue stub" that directs the agent to fetch its full implementation from a remote repository at https://raw.githubusercontent.com/nexu-io/open-design/main/skills/design-brief/SKILL.md. This source is not a recognized trusted organization or well-known service, posing a risk when downloading and executing third-party code.
[COMMAND_EXECUTION]: The skill includes a "Decision tree" section containing shell commands (e.g., grep, npx skills add) intended for the agent to execute. These commands are used to detect current installation status and perform the installation of the external skill content without manual user verification of the downloaded code.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted user input (design briefs) to generate output files, creating a potential surface for indirect injection.
Ingestion points: Ingests design briefs provided by users in either plain language or I-Lang format.
Boundary markers: The provided instructions lack explicit delimiters or safety instructions to prevent the agent from following malicious commands that might be embedded within a design brief.
Capability inventory: The skill (once installed) is authorized to write local files, specifically DESIGN.md and brief-preview.html.
Sanitization: There is no documentation or evidence of input validation or escaping for the user-supplied content before it is interpolated into the generated files.

design-brief