The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/inspect_pptx.py uses subprocess.run to execute the system utility fc-list. This is used to query the system for available font families to verify PPTX rendering compatibility.
[DYNAMIC_EXECUTION]: The script scripts/merge_images_to_pptx.py executes a local Node.js script (scripts/merge_images_to_pptx.mjs) by spawning a subprocess. The execution path for the Node binary and its modules is determined by the environment variables CODEX_PRIMARY_RUNTIME_NODE and CODEX_PRIMARY_RUNTIME_NODE_MODULES.
[DYNAMIC_EXECUTION]: The JavaScript file scripts/merge_images_to_pptx.mjs uses dynamic import and createRequire to load the @oai/artifact-tool at runtime based on environment-defined paths.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process and analyze external "benchmark" documents, project data files, and industry references provided by the user.
Ingestion points: Files like project-truth.md and user-supplied PDF/PPT reference files identified in benchmark-selection.md are ingested into the agent context.
Boundary markers: SKILL.md instructs the agent to categorize data into FACT, JUDGMENT, PROPOSAL, and UNKNOWN tags, which provides some structural separation between data and instructions.
Capability inventory: The skill possesses the ability to write files, generate images via external tools, and execute shell commands via the included Python and Node.js scripts.
Sanitization: There is no explicit sanitization or escaping mentioned for the external content before it is processed by the agent's reasoning logic.
[EXTERNAL_DOWNLOADS]: The documentation in docs/INSTALLATION.md provides a link to download the skill source code from the author's GitHub repository (github.com/dacnay816y62-hub/FANTASY-ppt-1.git).

ppt-image-deck