Open-source LLM knowledge platform: turn raw documents into a queryable RAG, an autonomous reasoning agent, and a self-maintaining Wiki.
Go agentagenticaichatbotdsh-pluginembeddingsevaluationgenerative-aigolangknowledge-basellmmulti-tenantollamaopenaiquestion-answeringragrerankingsemantic-searchvector-searchwiki
noVNC regression
Apply shared infrastructure PR #4141, then run npm run test:e2e -- e2e/novnc.spec.ts. The regression emulates a never-resolving VideoDecoder capability probe: upstream's static noVNC import blocks chat startup, while the lazy import leaves chat available. The browser harness and CI proposal are reviewed separately in #4141.