The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted content from external URLs and local audio/text files to generate articles and scripts. While it includes a sophisticated 'quality-review' loop, it lacks explicit boundary markers (e.g., XML tags or delimiters) or 'ignore' instructions when interpolating external content into prompts for the writing engine, creating a surface for indirect prompt injection.
[COMMAND_EXECUTION]: The workflow relies on multiple shell commands to orchestrate its pipeline, including curl for API requests, jq for JSON processing, base64 for image data handling, and local CLI tools like listenhub and coli for AI tasks. These commands are integral to the stated purpose of the skill and are restricted to managing content packages in the working directory.
[EXTERNAL_DOWNLOADS]: The skill connects to api.marswave.ai to perform content extraction and interacts with user-specified URLs to download media for transcription. These network operations are functional requirements for the content creation process.

creator