The Agent Skills Directory

[SAFE]: The skill demonstrates a strong security posture by including specific 'Self-Containment Rules' (Step 5) and a 'Design Contract' (Step 4) that strictly forbid the use of external font imports, CDN scripts, or any network-based fetching (fetch/XMLHttpRequest) in the generated HTML output.
[PROMPT_INJECTION]: The skill has a surface for Indirect Prompt Injection because it is designed to ingest and process untrusted data from a user's codebase. • Ingestion points: SKILL.md Step 1 (Codebase Scanning Protocol) instructs the agent to read various files including configuration, source code, and environment examples. • Boundary markers: No explicit boundary markers or 'ignore' instructions are provided to separate codebase content from the agent's primary instructions. • Capability inventory: The agent's capabilities are limited to generating a single HTML file; it is not instructed to execute commands, modify the file system (outside the output), or perform network operations. • Sanitization: While the skill extracts code snippets (Step 6), it does not explicitly define sanitization requirements, though the lack of execution capabilities limits the impact of this omission.

codebase-artifact