The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and apply 'voice profiles' from external YAML files located in the project directory (.aiwg/voices/) or the user's home configuration directory (~/.config/aiwg/voices/). These files directly influence the agent's output behavior, vocabulary choices, and narrative perspective.
Ingestion points: The voice_loader.py script reads and parses profile data from local files; the SKILL.md instructions guide the agent to load and apply these profiles dynamically.
Boundary markers: The skill instructions lack explicit delimiters or 'ignore' warnings for the profile content, which could allow instructions embedded within a voice profile to be misinterpreted as agent directives.
Capability inventory: The agent is directed to perform vocabulary replacement, tone calibration, and structural content adjustment based on data ingested from these external profiles.
Sanitization: The voice_loader.py script includes a validate_profile function that checks for required fields and numeric ranges for tone dimensions, providing structural validation but not semantic protection against malicious instructions.
[COMMAND_EXECUTION]: The skill relies on the execution of bundled Python scripts (scripts/voice_loader.py and scripts/voice_analyzer.py) to process and validate the voice configuration files at runtime.

voice-apply