sx4im

sx4im/skillcheck

A/B test agent skills with blind grading + bootstrap CIs - does your SKILL.md actually improve task performance?

26 stars 3 forks 26 watchers TypeScript MIT License
ab-testing agent-skill ai-agents benchmark claude-code claude-skill cli codex-skills developer-tools evals llm llm-eval nodejs prompt-engineering testing typescript
View on GitHub Website
2 Open Issues Need Help Last updated: Sep 11, 2026

Open Issues Need Help

View All on GitHub
feat(dashboard): migrate hosted proxy upstream from NVIDIA NIM to Groq 11 days ago
enhancement help wanted
sx4im/skillcheck
26

A/B test agent skills with blind grading + bootstrap CIs - does your SKILL.md actually improve task performance?

TypeScript
#ab-testing#agent-skill#ai-agents#benchmark#claude-code#claude-skill#cli#codex-skills#developer-tools#evals#llm#llm-eval#nodejs#prompt-engineering#testing#typescript
refactor(adapters): consolidate NvidiaNimClient into OpenAiCompatClient 12 days ago
enhancement good first issue
sx4im/skillcheck
26

A/B test agent skills with blind grading + bootstrap CIs - does your SKILL.md actually improve task performance?

TypeScript
#ab-testing#agent-skill#ai-agents#benchmark#claude-code#claude-skill#cli#codex-skills#developer-tools#evals#llm#llm-eval#nodejs#prompt-engineering#testing#typescript