The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to execute mkdir for managing output directories and rg for security validation of generated HTML content. These operations are scoped and intended for local file management.
[DATA_EXFILTRATION]: The skill implements a strict "no external resources" policy. It requires the agent to run a rg (ripgrep) command to verify that the generated HTML does not include CDN links, remote scripts, or network request functions like fetch or import that could leak data.
[PROMPT_INJECTION]: The skill processes data from user prompts and repository files to populate HTML templates, creating a surface for indirect prompt injection. However, the risk is mitigated by the skill's enforcement of self-contained, offline-only artifacts and its validation checklist.
[SAFE]: No evidence of credential theft, obfuscated code, persistence mechanisms, or unauthorized privilege escalation was found. The skill follows developer best practices for safe, isolated artifact generation.

html-artifacts