The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill directs the user to install a plugin from a third-party repository (nexu-io/codex-slides) and references a related SDK package (@codex-slides/sdk). These sources are not recognized as trusted organizations or well-known services within the defined scope.
[COMMAND_EXECUTION]: The installation process involves executing shell commands (codex plugin marketplace add, codex plugin add) that install and register external tools on the local environment.
[INDIRECT_PROMPT_INJECTION]: The skill's primary function involves ingesting untrusted data from local files to generate slide decks, creating a surface for prompt injection.
Ingestion points: The skill reads files from user-defined paths such as ./pitch-materials/, ./docs/, and specific files like sales-data.json to extract content for outlines.
Boundary markers: There are no explicit instructions or delimiters defined to ensure the agent ignores potentially malicious instructions embedded within the source markdown, PDF, or JSON files.
Capability inventory: The skill utilizes several capabilities including reading the filesystem, executing MCP tools for slide rendering, and writing files during the PPTX and PDF export process.
Sanitization: The instructions do not define any sanitization, filtering, or validation steps for the content extracted from external source files before it is processed by the LLM.

codex-slides-ai-presentation