The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or unauthorized access attempts were detected. The skill operates entirely within the expected parameters of the vendor's platform.
[COMMAND_EXECUTION]: Employs the infsh CLI tool to manage AI workloads. The skill's execution environment is constrained to this specific toolset via the allowed-tools configuration in the YAML frontmatter.
[EXTERNAL_DOWNLOADS]: Retrieves configuration and modular extensions from the vendor's official GitHub repository (inference-sh/skills). These references are standard for the platform's architecture and originate from the author's own infrastructure.
[PROMPT_INJECTION]: The skill contains an indirect prompt injection surface as it is designed to process untrusted user documents into podcast scripts.
Ingestion points: User-supplied document content used as input for the openrouter/claude-sonnet-45 model in SKILL.md.
Boundary markers: Not present in the provided script-writing templates.
Capability inventory: Execution of LLMs, audio synthesis models, and media merging tools via the infsh CLI as defined in SKILL.md.
Sanitization: Input content is passed directly to the AI models without pre-processing or explicit boundary isolation.

ai-podcast-creation