The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user content from the content field to generate infographic specifications. This creates a surface for potential indirect prompt injection where instructions hidden in the data could affect the agent's output.
Ingestion points: content field in the input schema defined in SKILL.md.
Boundary markers: None present; the skill lacks specific markers to separate data from system instructions.
Capability inventory: Generation of structured data (JSON/YAML) and renderable HTML/CSS as seen in the Step 6 and Step 7 workflows of SKILL.md.
Sanitization: No sanitization or input validation logic is described in the workflow instructions.
[DYNAMIC_EXECUTION]: The skill generates self-contained HTML and CSS code blocks based on user-supplied data and selected layouts. While intended for visualization, this represents a form of script generation that could be misused if the output is rendered in a sensitive environment without sanitization.
[EXTERNAL_DOWNLOADS]: The skill references Vercel's Satori library on GitHub and the author's own content-pipeline repository as resources for rendering the generated specifications.

infographic-generator