The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user data to populate an HTML presentation template, which is then written to the file system. This creates a surface for indirect prompt injection where malicious data could attempt to influence the agent's behavior.
Ingestion points: User-provided topic, narrative details, and financial data (SKILL.md).
Boundary markers: Absent; the skill lacks specific instructions to delimit user input or warn the agent about potential instructions within the data.
Capability inventory: The skill has the capability to write to the file system (fs:write), which is used to generate the final HTML artifact.
Sanitization: Absent; there is no explicit requirement for the agent to sanitize or escape user-provided content before placing it into the HTML structure.
[EXTERNAL_DOWNLOADS]: The skill's HTML template fetches font assets from a well-known external service.
Evidence: @import of Google Fonts from fonts.googleapis.com in example.html.
Context: This is a standard practice for loading web fonts and does not represent a security risk.

huashu-sparkline-arc