The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes several Python and shell scripts designed for project management and slide generation, such as scripts/scaffold_deck.py for template creation, scripts/generate_image_deck.py for API interaction, scripts/index_image_sources.py for asset discovery, and scripts/install_local.sh for environment setup.
[EXTERNAL_DOWNLOADS]: The SKILL.md instructions guide the agent to install additional capabilities from the author's monorepo using an external skill installation script.
[DATA_EXFILTRATION]: The scripts/generate_image_deck.py script communicates with the official Google Gemini API at generativelanguage.googleapis.com to render slide visuals. This is a well-known service and is documented as a core feature.
[PROMPT_INJECTION]: The skill processes user-supplied markdown and JSON slide specifications to generate prompts for an AI image model, which constitutes a vulnerability surface for indirect prompt injection.
Ingestion points: Slide text and visual cues are read from local files by scripts/generate_image_deck.py.
Boundary markers: While structural delimiters are used for parsing, there are no explicit instructions to ignore potentially malicious commands embedded within slide content.
Capability inventory: The skill has permission to write local files and initiate network requests to an external API.
Sanitization: No input filtering or validation is performed on the slide content before it is interpolated into the final API prompts.

slides