The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local shell commands to facilitate rendering.
Evidence: It starts a local HTTP server using python3 -m http.server 8765 to serve the generated render.html file to Playwright.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it handles untrusted data that is later rendered in a browser context.
Ingestion points: The skill reads card copy and text content from the 'planner' output and interpolates it into HTML templates.
Boundary markers: The skill uses {{placeholder}} syntax for interpolation but does not describe using delimiters or instructions to ignore embedded commands within the data.
Capability inventory: The skill has the ability to execute shell commands (python3 -m http.server) and control a browser via Playwright MCP (including browser_evaluate).
Sanitization: There is no explicit mention of HTML escaping, sanitization, or validation of the input text before it is placed into the render.html file. Malicious input could potentially execute script in the rendering browser.
[EXTERNAL_DOWNLOADS]: The skill references external assets required for rendering.
Evidence: It fetches the 'Pretendard' font stylesheet from jsDelivr (https://cdn.jsdelivr.net/gh/orioncactus/pretendard/dist/web/static/pretendard.css). As this targets a well-known service for standard assets, it is documented neutrally.

storyteller-renderer