The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides several Python scripts (visualize_skill.py, profile_store.py, validate_skill.py) designed for local execution by the agent to analyze skill directories and manage user preferences.
[INDIRECT_PROMPT_INJECTION]: The skill performs static analysis on user-supplied skill files, which presents a surface for indirect prompt injection if those files contain malicious instructions.
Ingestion points: Local SKILL.md files and associated references are read by visualize_skill.py to extract logic flows.
Boundary markers: The tool uses regex-based extraction to identify specific Markdown sections but does not implement explicit boundary markers for the data it processes.
Capability inventory: The skill performs file reading and report generation (Markdown, HTML, JSON) but explicitly avoids executing instructions found within the analyzed skills.
Sanitization: The skill utilizes yaml.safe_load for manifest parsing and performs classification of text as data rather than executable instructions.

lov-skill-visualizer