Add artemiskit-cli skill

[code-sensei]

Summary

Adds the artemiskit-cli skill for LLM evaluation and security testing.

ArtemisKit is an open-source CLI toolkit that helps developers:

• Test LLM outputs with scenario-based evaluation (YAML-driven quality testing)
• Secure LLMs via red teaming (prompt injection, jailbreaks, data extraction, PII disclosure)
• Stress test LLM endpoints (latency p50/p95/p99, throughput, token usage, cost estimation)
• Compare evaluation runs for regression detection
• Generate interactive HTML reports and JSON manifests

Commands

Command	Purpose
akit run	Execute scenario-based evaluations
akit redteam	Security red team testing
akit stress	Load and stress testing
akit report	Generate/regenerate reports
akit history	View run history
akit compare	Compare two evaluation runs
akit baseline	Manage baselines for regression testing
akit validate	Validate scenario files
akit init	Initialize configuration

Provider Support

• OpenAI (GPT-4, GPT-4o, etc.)
• Anthropic (Claude 3.5, Claude 4, etc.)
• Azure OpenAI
• Vercel AI SDK
• OpenAI-compatible APIs (Ollama, vLLM, LM Studio)

Skill Structure

skills/artemiskit-cli/
├── SKILL.md              # Main skill file
└── references/
    ├── commands.md       # CLI command reference
    ├── providers.md      # Provider configuration
    └── scenarios.md      # Scenario format documentation

Links

• Repository: https://github.com/code-sensei/artemiskit
• npm package: @artemiskit/cli
• Documentation: https://artemiskit.vercel.app

Test Plan

• Skill follows SKILL.md format with YAML frontmatter
• References use progressive disclosure pattern
• All commands verified against CLI implementation (--help output)
• Installed and tested via npx skills install code-sensei/artemiskit-cli-skill