Case study
Agent Sandbox
Browser-local LLM and RAG playground
Problem
TODO: Describe the friction of testing LLMs and RAG pipelines without standing up backends or leaking API keys.
Approach
TODO: Outline the zero-backend architecture, IndexedDB storage, and multi-provider compare mode.
Outcome
TODO: Add adoption metrics, demo usage, or qualitative feedback once available.
This page is a scaffold — replace TODO sections in src/content/case-studies.ts with final narrative and metrics.