The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill possesses a data ingestion surface by reading and extracting text from untrusted .pptx files using the extract_text and unpack actions of the Presentation tool.
Ingestion points: Content is ingested from user-provided .pptx files in SKILL.md workflows.
Boundary markers: The skill does not explicitly instruct the agent to use delimiters or ignore instructions embedded within the slide content.
Capability inventory: The skill has access to powerful tools including Bash, Write, Edit, and Presentation, which could be targeted if an agent follows instructions found inside a malicious presentation.
Sanitization: No explicit sanitization or validation of the extracted slide text is mentioned before processing.
[COMMAND_EXECUTION]: The skill utilizes the Bash tool to run commands like pptxgenjs and python scripts/thumbnail.py. These are legitimate uses for the skill's stated purpose of generating and processing presentation files.

pptx