The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes local shell commands to manage its workflow, including creating temporary directories, synchronizing template files, and launching a Next.js development server (bun dev --port 3333) to provide a live preview of the generated carousel. These operations are restricted to the local environment and are necessary for the skill's stated purpose.
[EXTERNAL_DOWNLOADS]: The skill's template includes a package.json file that specifies standard dependencies such as next, react, html-to-image, and jspdf. These are fetched from official package registries during the initial setup (bun install). These sources are well-known and standard for web development.
[PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection (Category 8) because it processes untrusted user-supplied text to generate slide content.
Ingestion points: Untrusted data enters the agent context through user-provided text posts or Markdown files referenced in Step 1 of the workflow.
Boundary markers: The instructions do not specify the use of boundary markers or 'ignore' instructions when the agent interpolates user text into the src/slides.ts file.
Capability inventory: The skill possesses the capability to write files to the local file system and execute shell commands (cp, rsync, bun dev).
Sanitization: No specific sanitization or escaping logic is defined for the user input; the skill relies on the agent's internal logic to correctly format the data into a TypeScript array structure.

threads-carousel