The Agent Skills Directory

[COMMAND_EXECUTION]: The skill orchestrates its workflow using various shell commands and local CLI tools.
Uses mkdir, cat, and rm for state management and output preparation in the working directory and .listenhub/creator/.
Employs jq for parsing data from API responses and configuration files.
Executes listenhub and coli CLI tools to perform image generation, text-to-speech, and transcription tasks.
[EXTERNAL_DOWNLOADS]: Fetches external resources required for content processing.
Downloads user-provided media URLs to a temporary directory using curl for subsequent transcription.
Uses a remote API to extract content from arbitrary web URLs provided by the user.
[DATA_EXFILTRATION]: Transmits processed content and prompts to external services.
Sends user-provided material and generated instructions to api.marswave.ai for content extraction and media generation.
These interactions are documented as part of the vendor's own infrastructure and are necessary for the skill's functionality.
[PROMPT_INJECTION]: The skill possesses a vulnerability surface for indirect prompt injection due to its handling of untrusted data.
Ingestion points: External data enters the agent's context through web content extraction (Step 5 of the pipeline) and the reading of local audio/text files.
Boundary markers: The prompt templates for writing articles and scripts do not use explicit delimiters or "ignore previous instructions" warnings when interpolating the extracted material.
Capability inventory: Across its scripts, the skill has the ability to execute shell commands, perform network operations, and write to the local file system.
Sanitization: There is no evidence of filtering, escaping, or validation applied to the external content before it is processed by the AI for content generation.

creator