The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The pptxgenjs.md reference documentation specifies that the addImage function can fetch images from external remote URLs (e.g., slide.addImage({ path: 'https://example.com/image.jpg', ... })). This is a standard feature for presentation software but involves outbound network requests to retrieve assets.
[COMMAND_EXECUTION]: The create_presentation tool accepts a custom_code parameter which executes arbitrary JavaScript within the scope of the PptxGenJS library. This dynamic code execution is the primary mechanism for generating complex slide layouts and visual elements.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes data from external sources.
Ingestion points: The analyze_presentation and update_slide_content tools read text and structure from user-provided .pptx files. Additionally, the addImage tool can fetch content from remote URLs.
Boundary markers: No specific delimiters or safety instructions are defined to separate untrusted text found in slides from the agent's instructions.
Capability inventory: The skill possesses capabilities for reading/writing files and making network requests to fetch images.
Sanitization: The instructions do not specify any sanitization or validation of the text content extracted from presentations before it is used in subsequent reasoning steps.

powerpoint-presentations