The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill documentation references and depends on an external tool slides-grab (hosted at https://github.com/vkehfdl1/slides-grab). This repository belongs to an unverified individual user rather than a trusted organization or well-known service provider.
[COMMAND_EXECUTION]: The skill's instructions require the agent to execute various shell commands using the slides-grab CLI, such as build-viewer, validate, edit, convert, and pdf. Running unvetted third-party binaries can lead to arbitrary code execution if the tool is malicious or compromised.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection (Category 8) due to its core workflow of processing untrusted external data.
Ingestion points: The agent is instructed to ingest product briefs, documents, and other source materials (e.g., in SKILL.md Steps 1 and 2) to generate slide content.
Boundary markers: The instructions lack specific guidance on using delimiters (like XML tags or triple backticks) or instructions to ignore potential commands embedded in the source materials.
Capability inventory: The skill possesses the ability to read and write files and execute shell commands via the slides-grab tool.
Sanitization: There is no evidence of sanitization or validation logic to filter out malicious instructions that might be present in the user-provided briefs or documentation.

presentation-builder