The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructions require the agent to read and process untrusted HTML files to determine navigation logic and slide structure. Malicious content within these files could attempt to influence the agent's behavior.
Ingestion points: HTML slide decks are accessed using the read_file tool (SKILL.md).
Boundary markers: There are no specific delimiters or warnings used to prevent the agent from following instructions embedded in the slide content.
Capability inventory: The skill utilizes the gen_pptx tool which can execute JavaScript and the show_to_user tool.
Sanitization: The instructions do not specify any sanitization or validation of the HTML content before it is processed by the agent.
[DYNAMIC_EXECUTION]: The gen_pptx tool accepts a showJs parameter, which is used to execute arbitrary JavaScript strings within an iframe context for the purpose of slide navigation. The skill provides templates for generating these scripts based on the ingested content.
[EXTERNAL_DOWNLOADS]: The skill can be configured to fetch assets from Google Fonts via the googleFontImports parameter. This facilitates proper rendering of slide typography from a well-known service.

export-as-pptx-editable