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
import { nextTick, onMounted, ref, watch, type Ref } from 'vue'
/**
* One-shot attention motion when generated files first land on a live
* answer. Historical messages already have artifacts at mount and stay still.
*/
export function useArtifactArriveMotion(artifactCount: Ref<number>) {
const artifactArrived = ref(false)
let armed = false
onMounted(() => {
requestAnimationFrame(() => {
armed = true
})
})
watch(artifactCount, (count, previous) => {
if (!armed || count <= 0 || (previous ?? 0) > 0) return
artifactArrived.value = false
nextTick(() => {
artifactArrived.value = true
})
})
function onArtifactArriveEnd(event?: AnimationEvent) {
if (event?.animationName && event.animationName !== 'artifact-toolbar-arrive') return
artifactArrived.value = false
}
return { artifactArrived, onArtifactArriveEnd }
}