AI, MCP & RAGNotable
automation-kb-studio
An LLM evaluation harness plus a browser-based 'studio' for authoring knowledge bases and comparing coverage. Probes model/RAG backends with fabricated test cases to expose gaps; ships vendor-neutral core with OpenAI- and Anthropic-compatible providers, a CLI, and a reporter.
0
stars
0
forks
TypeScript
language
active
status
What's inside
@evaluator/core, provider-openai, provider-anthropic, reporter (JSONL/console), cli (health/schema/coverage), studio (browser KB authoring); folder-first KB structure
Category
AI, MCP & RAG — one of 7 repos in this group.