The Agent Skills Directory

[COMMAND_EXECUTION]: The skill invokes local Python scripts (scripts/content/analyze_brand_voice.py, generate_voice_guide.py, and check_brand_alignment.py) to process content and generate reports. These scripts are executed within a local virtual environment (.venv\Scripts\python.exe), following standard development practices for packaged skills.
[PROMPT_INJECTION]: The skill identifies a surface for indirect prompt injection by ingesting and processing untrusted content from references/examples/ and output/content/draft.md during voice audits and alignment checks. \n
Ingestion points: references/examples/ (via analysis script) and output/content/draft.md (via alignment check script). \n
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the skill definition. \n
Capability inventory: The agent has access to Read, Write, and Bash tools, which could potentially be targeted by instructions hidden in processed data. \n
Sanitization: The instruction set does not specify validation or sanitization routines for the content before it is processed by the analysis scripts.

brand-voice-manager