The Agent Skills Directory

[SAFE]: The skill follows secure development practices for AI agent extensions, utilizing structured data (JSON) for communication between its component scripts.
[SAFE]: The provided Python scripts (funnel_gap_analyzer.py, root_cause_ranker.py, and corrective_action_prioritizer.py) perform deterministic data processing and scoring. They do not initiate external network connections, access sensitive system files, or use unsafe functions like eval() or exec().
[SAFE]: Command execution is limited to the skill's own local Python scripts with pre-defined argument patterns for processing input files.
[SAFE]: No obfuscation, hidden instructions, or credential exposure was found across the SKILL.md, references, or scripts.
[SAFE]: Indirect prompt injection risk is minimal as the skill processes structured performance data and uses web searches only for standard industry benchmarks.

performance-diagnosis