The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on the execution of shell commands via the uvx tool to perform its primary functions, including presentation rendering, preflight validation, and quality assurance (e.g., uvx --from agent-slides slides render ...).
[EXTERNAL_DOWNLOADS]: The skill uses uvx to fetch the agent-slides package from a remote registry (PyPI) during execution. This package is not provided by a recognized trusted organization, which introduces risks associated with third-party software dependencies.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user content, such as presentation briefs and outlines, to generate a slide deck. This creates a potential attack surface where malicious instructions hidden in the input could influence the agent's planning or content generation logic.
Ingestion points: User-provided briefs, topics, and outlines ingested in the intake step.
Boundary markers: Absent; there are no instructions to the agent to treat user input as untrusted or to use specific delimiters.
Capability inventory: The skill has the ability to write local files (slides.json) and execute shell commands (uvx).
Sanitization: No specific sanitization, validation, or filtering of the user-provided content is performed before it is used to generate the presentation manifest.

slides-build