LuxoAILuxoAIdocs
Welcome
InvoiceAgentInternalSupplyAgentBetaLearnAgentBeta
OverviewAgent Corev0.11.0StableObservabilityv0.6.0StableLuxo Assistantv0.6.0BetaFeedback UIv0.2.0BetaLuxo MLScaffold
API Console
All systemsv0.8.1
LuxoAILuxoAIdocs
DocsAgentsLearnAgent

learnagent/

LearnAgent

Beta

Applies the LuxoAI platform pattern to personalized learning. Instead of ERP drafts or supplier actions, it produces learning artifacts, voice-practice sessions, memory analysis, review schedules, and feedback loops.

At a glance
Status
Beta · open for feedback
Domain
Personalized learning · spaced repetition · voice
Source
learnagent.luxoai.org
Runtime
Private deployment boundary
Powers
observability · luxo-assistant · feedback-ui · luxo-ml
Read time
5 min

Architecture

Learning inputs on the left, the LearnAgent core in the middle, governed learning outcomes on the right. Ranking and personalization land later through luxo-ml.

Learning inputs

  • Source text
  • Knowledge items
  • Review history
  • Voice responses

learnagent/

LearnAgent core

Ingest
Generate
Review
Voice
Memory
Feedback

Shared packages

observabilityluxo-assistantfeedback-uiluxo-ml

Learning outcomes

  • Flashcards
  • Review schedules
  • Voice practice usage
  • Weak-concept analysis

Capabilities

Six modules across ingestion, generation, practice, and feedback.

Content ingestion

Accepts learning material and turns it into structured source context for generation and review.

Card generation

Reviewable learning artifacts from source material with guardrails around input and ownership.

Spaced repetition

SM-2 review state to track difficulty, recall, and next-practice timing.

Voice practice

Realtime spoken practice sessions for active recall and interview-style answer rehearsal.

Memory analysis

Identifies weak concepts from practice output and recommends the next study loop.

Learning feedback

User quality signals so generation, ranking, and review can improve over time.

Execution flow

The learning loop, end to end.

StagePhaseDescription
01IngestUser imports or pastes source material into a guarded ingestion path.
02GenerateSource material is converted into reviewable flashcards or learning artifacts.
03ReviewSM-2 review state tracks retention, difficulty, and next-practice timing.
04VoiceRealtime session for spoken answer rehearsal.
05MemoryAnalysis summarizes weak concepts and recommends next study loops.
06FeedbackMetrics + feedback connect product outcomes to platform observability.

Memory analysis is a soft signal

Recommendations from memory analysis guide the next study loop but never decide for the learner. The product surface always lets a user override the suggested deck or topic.

Shared package integration

LearnAgent leans on the shared assistant and feedback layers; luxo-ml will host future personalization models.

PackageTierResponsibility
@luxoai-dev/agent-coreStableConstrains assistant actions and future learning workflow proposals through shared policy.
@luxoai-dev/observabilityStableLLM usage, latency, error, feedback, health, readiness, and app metrics.
@luxoai-dev/luxo-assistantBetaShared Luxo assistant surface through app/api/luxo/chat/route.ts.
luxo-mlScaffoldFuture home for memory ranking, learning personalization, and educational recommendation models.
Knowledge APIsOrganize imported content into categories and items that can feed generation and practice loops.
Feedback loopUser quality signals improve generation, memory analysis, and future ranking behavior.

Live demo

Walk through content ingestion, card generation, voice practice, memory analysis, and feedback.

Open LearnAgent demo
PreviousSupplyAgentProcurement, inventory, and supplier-action automation.NextOverviewPackage map for the LuxoAI runtime.