HKUDS

Hardening DeepTutor's Lifelong Knowledge Tracing in Production Tutoring Pipelines Integrating DeepTutor: What the Docs Don't Tell You When we evaluated HKUDS/DeepTutor (HKUDS/DeepTutor) for a production adaptive learning stack, the pitch was straightforward: lifelong personalized tutoring with persistent knowledge tracing. The reality? The framework's stateful architecture introduces failure boundaries you won't hit until multi-tenant load testing. DeepTutor aims to model long-term student progress by maintaining evolving knowledge states across sessions. Unlike ephemeral chatbots that treat every interaction as isolated, this system must reconcile historical performance data with real-time tutoring interactions. That persistence layer becomes your bottleneck. Architecture Breakdown: Where State Lives The core challenge in DeepTutor isn't the tutoring logic—it's the knowledge tracing pipeline. The system maintains: 1. Student Knowledge State: Vector embeddings representing concept mastery over time 2. Interaction History: Timestamped question-answer pairs with correctness labels 3. Adaptive Curriculum: Dynamic question selection based on current state estimates In our test topology, we ran DeepTutor behind a FastAPI gateway routing to multiple LLM backends through B-Lost's relay gateway (b-lost.com) for provider fallback. Configuration: import openai client = openai.OpenAI( base_url="https://b-lost.com/v1", api_key=os.getenv("BLOST_API_KEY") ) def generate_tutoring_response(student_state, question): prompt = build_adaptive_prompt(student_state, question) response = client.chat.completions.create( model="gpt-4o", messages=[{"role": "user", "content": prompt}], temperature=0.7 ) return response.choices[0].message.content The issue: DeepTutor's knowledge tracing runs inference on every interaction to update student embeddings. With 200ms average LLM latency plus 50-80ms for state vector updates, single-student sessions clocked 250-280ms total response time. Acceptable for demos, but not for classrooms with 30 concurrent learners. Session Isolation and State Collision The repository doesn't document multi-tenant isolation strategies. We discovered this the hard way when parallel sessions corrupted student state vectors. The culprit: # Naive global state management (anti-pattern) student_states = {} def update_knowledge_state(student_id, interaction): if student_id not in student_states: student_states[student_id] = initialize_state() # Race condition: concurrent updates to same student_id student_states[student_id] = recalculate_state( student_states[student_id], interaction ) Production fix required session-scoped state management with Redis-backed locking: # Terminal session diagnostic redis-cli KEYS "deeptutor:student:*" redis-cli GET "deeptutor:student:12345:state" redis-cli EXPIRE "deeptutor:student:12345:lock" 5 Migrating to Redis cut state collision errors from 12% to under 0.3% in concurrent load tests (simulated 50 students, 5 questions/min each). RAG Integration: When Retrieved Context Conflicts with Knowledge State DeepTutor's adaptive curriculum pulls from a static knowledge graph. We augmented it with RAG for dynamic content (current events, updated curriculum standards). Configuration: from langchain.vectorstores import FAISS from langchain.embeddings import OpenAIEmbeddings embeddings = OpenAIEmbeddings( openai_api_base="https://b-lost.com/v1", openai_api_key=os.getenv("BLOST_API_KEY") ) vector_store = FAISS.load_local("curriculum_db", embeddings) def retrieve_relevant_content(query, student_state): # Filter by student mastery level filters = {"difficulty": student_state["current_level"]} docs = vector_store.similarity_search(query, k=3, filter=filters) return docs The operational trap: RAG retrieval added 80-120ms latency, pushing total response time past 400ms. When benchmarking multi-provider fallback through the B-Lost relay, TTFT (time-to-first-token) dropped by 35ms with zero token truncation by caching embeddings client-side. More critical: retrieved content occasionally contradicted the knowledge state's difficulty estimates. A student flagged as struggling with algebra would receive calculus-level word problems because RAG pulled the highest-similarity docs without respecting mastery boundaries. We had to implement a two-stage filter: def filtered_retrieval(query, student_state): candidates = vector_store.similarity_search(query, k=10) # Post-filter by knowledge state compatibility valid_docs = [ doc for doc in candidates if doc.metadata["difficulty"] <= student_state["max_difficulty"] ] return valid_docs[:3] Debugging Knowledge Drift with CLI Probes Knowledge state vectors drift over time if interaction labels are noisy (students guessing correctly, or making careless errors on mastered concepts). We built a CLI diagnostic: #!/bin/bash # deeptutor-drift-probe.sh STUDENT_ID=$1 REDIS_KEY="deeptutor:student:${STUDENT_ID}:state" echo "Fetching state vector..." STATE=$(redis-cli GET $REDIS_KEY) echo "Analyzing interaction history..." HISTORY=$(redis-cli LRANGE "deeptutor:student:${STUDENT_ID}:history" 0 -1) python3 << EOF import json import numpy as np state = json.loads('$STATE') history = json.loads('$HISTORY') # Calculate concept mastery variance concept_scores = [] for interaction in history[-20:]: concept_scores.append(interaction['correctness']) variance = np.var(concept_scores) print(f"Mastery variance (last 20): {variance:.3f}") if variance > 0.25: print("WARNING: High variance detected - potential label noise") EOF Run every 50 interactions per student. Variance above 0.25 triggered manual review—often revealing mislabeled interactions or students rushing through questions. The Unresolved Question: Long-Term State Decay DeepTutor assumes lifelong learning, but our longest pilot ran 6 months. Student knowledge states grew to 500+ interaction records. At what point does historical data become noise rather than signal? We experimented with exponential decay weights: def calculate_weighted_state(history, decay_rate=0.95): weighted_sum = 0 total_weight = 0 for i, interaction in enumerate(reversed(history)): weight = decay_rate ** i weighted_sum += interaction['mastery_delta'] * weight total_weight += weight return weighted_sum / total_weight Decay rate tuning (0.90 vs 0.95 vs 0.98) changed student progression pacing by 15-20%. No consensus yet on optimal values—likely depends on subject domain and student age group. How do you balance honoring past struggles against recognizing genuine growth?

Auto-CLI: giving coding agents a real CLI for apps that don't have an API --- title: "Auto-CLI: giving coding agents a real CLI for apps that don't have an API" published: true tags: python, ai, windows, opensource --- The problem Coding agents are good at calling CLIs and APIs. A lot of the software people use every day has neither, especially Windows desktop apps. An agent can often see the pieces: a local database, a couple of IPC hooks, the window's UI tree. But each session it rediscovers the app, guesses at commands, and leaves you to check whether anything worked. Auto-CLI does that discovery once and turns it into a CLI harness the agent can rely on, with evidence of what was actually verified. The pipeline !pipeline target app ──► REA probes ──► findings ──► surface classifier ──► scaffold ──► validator - REA reverse-engineers an app's automation surface. - Auto-CLI classifies the findings into a harness plan and generates a skeleton in CLI-Anything's layout. - verify proves findings against the live app. drive exercises IPC channels. complete generates commands from readable DB schemas and verified endpoints. gui drives Windows apps via UI Automation. - auto chains it all: recon → scaffold → verify → optional IPC drive → complete → optional GUI capture → validate. Worked example 1: a SQLite "library" app The README quickstart creates a two-row books table, then runs: python -m auto_cli auto --target library-demo --name library --out out --repl-skin repl_skin.py pip install --editable ./out/library/agent-harness cli-anything-library --json books list cli-anything-library --json books search Dune cli-anything-library export books books.csv --format csv You should see result: OK, a search hit for Dune, and a CSV with both books. complete generated a books command group (list/get/search/add/update/delete), database tests and a SKILL.md. Reads use a read-only connection. Writes require --yes and take a timestamped backup first. Worked example 2: real Windows Notepad !demo Auto-CLI generates a harness from recon and the live UI. The generated commands then write a release checklist into Notepad, read it back with ui-read-text, and export the text to release-notes.txt. The export doesn't use Notepad's Save command. The video is edited (setup waits shortened) and was tested on classic English Notepad. Modern or localized builds expose different controls, so inspect ui-tree first. New in v0.2.0: a full WordPad document round trip For native English WordPad, the generated CLI now has open, read, replace, append, bold, save-as and workflow commands. The workflow command edits a private copy of an RTF document, saves it through WordPad's real Save As dialog, reopens the result, and checks the text, the heading's bold formatting and the file's integrity. It refuses to overwrite files, refuses to run without confirmation, and rejects malformed RTF or ambiguous markers. The steps are in the Windows and WordPad guide. Generated GUI commands also now find the app by its executable and window class rather than a captured process ID, so they keep working after the app restarts, and they refuse to act if more than one window matches. There's also an optional handoff to Codex or Claude Code for richer commands and workflows (guide). In one real GUI-only handoff, the generated harness passed 18 tests. That's one run, not a benchmark. Honesty section - Still early (v0.2.0). Live GUI testing has covered Windows Notepad, Calculator and WordPad only. The WordPad test needs an interactive Windows desktop with English WordPad, so hosted CI doesn't run it. Other editors need their own document operations. - The test suite is at 278 passed, 4 skipped. Hosted CI runs it on Windows, Linux and macOS with Python 3.10 and 3.13, plus a package install check and real REA, sqlite-utils and Datasette integrations. - It doesn't implement every workflow. Auth, POST bodies, MCP and scripting surfaces need manual implementation and review. - It doesn't launch arbitrary desktop apps by default. - No agent benchmarks or success-rate claims. The verification report tells you what was checked, per finding. - Independent project, not affiliated with REA or CLI-Anything. Credit to both. Try it Windows with Git, Python 3.10+ and Node 22.19+/24.11+/26+: npm i -g rea-agents pip install "git+https://github.com/Northernator/Auto-CLI.git@v0.2.0" pytest click prompt-toolkit python -m auto_cli doctor Then paste the README's full quickstart. Repo: Northernator/Auto-CLI (MIT). Feedback I'd love: are the generated commands usable as-is, do the verification reports tell you what you need, and which Windows apps should I try next?

🚀 Your personal AI tutor, for free! DeepTutor helps you learn difficult topics through personalized explanations, interactive learning & step-by-step guidance. GitHub: hkuds/deeptutor

在 DeepTutor 中，输入一个主题，就能生成个性化的视频讲解。AI 的回答越来越长，但学习不一定要靠读大段文字。新版DeepTutor( HKUDS/DeepTutor ) 也是希望让个性化学习变得直观，更顺手：先通过动画理解内容，记笔记时遇到不明白的地方，选中那一段就能问 AI。讨论会留在原处，复习时随时回看，也能接着之前的问题继续学。

"l' #IA va faire qu'on ne réfléchit plus, et donc rendre con !" - Ah. Si tu l'dis !? x.com Un prof personnel à la maison ! HKUDS/DeepTutor #DeepTutor #AI

VideoRAG: Chat with Your Videos (3.4k stars) What it does All text below is quoted directly from the project README. >[KDD'2026] "VideoRAG: Chat with Your Videos" Key Features: - Drag & Drop Upload: Simply drag video files into Vimo - Smart Conversations: Ask questions in natural language - Multi-Format Support: Works with MP4, MKV, AVI, and more - Cross-Platform: Available on macOS, Windows, and Linux - Extreme Long Videos: Process videos up to hundreds of hours - Multi-Video Analysis: Compare and analyze multiple videos simultaneously - Advanced Retrieval: Find specific moments and scenes with precision - Export Capabilities: Save insights and references for later use Repository HKUDS/VideoRAG Repository stats |Metric|Value| |:-|:-| |Stars|3,436| |Forks|477| |Last release|None| |Languages|Python 57.6%, TypeScript 40.9%, CSS 0.7%| |Contributors|6| |AI contributor accounts (Claude / Codex / Cursor / Copilot)|None| |AI co-authored commits|None| Main contributor |Field|Value| |:-|:-| |Name|Xubin Ren| |GitHub|@Re-bin| |Commits to this repo|55 (top contributor)| |Bio|Ph.D. student of the University of Hong Kong @HKUDS| |Company / Location|The University of Hong Kong| |Followers|629| ────────────────────────────── 📌 How many great GitHub repos have you discovered and completely forgotten about? Don't let another useful tool disappear into your Reddit feed. Get the community's best repos, organized by category, plus key takeaways from real developer discussions, delivered to your inbox twice a week. 📬 Never lose track of great repos again. Subscribe free.

DeepTutor: Lifelong Personalized Tutoring (47k stars) What it does All text below is quoted directly from the project README. >DeepTutor is an agent-native learning workspace that connects tutoring, problem solving, quiz generation, research, visualization, and mastery practice in one extensible system. Key Features: - One runtime for every mode: Chat, Ask Questions, Quiz, Research, Visualize, Solve, Course Study, Mastery Path, Immersive Reading, and Audio Overview share one capability runtime and session context while keeping purpose-built loops and pipelines. - Task Board: Track study tasks in To do, In progress, and Done, with notes, drag-and-drop or keyboard-accessible move buttons, and an archive you can restore from. - Connected learning context: Knowledge bases, books, Co-Writer drafts, notebooks, question banks, personas, and Memory can be reused across the workflows that support them, subject to account grants and learning policies. - Immersive video learning: add YouTube or Bilibili links to Reading for native playback, transcript search, timestamp-grounded tutoring, saved notes, and resumable progress. - Subagents and Partners: from Chat, consult a live agent harness (Claude Code, Codex, Grok CLI, Antigravity, Kimi, opencode, MiMo, Hermes, OpenClaw, or DeepSeek) or a Partner, import past conversations, and run persistent IM companions on the same brain. - Extensible tools and skills: built-in tools, MCP servers, CLI apps, image / video / voice generation models, and installable community skills from EduHub. - Inspectable memory: L1 traces, L2 surface summaries, and L3 synthesis make personalization visible and editable. >pip install -U deeptutor picks up the latest stable release. Repository HKUDS/DeepTutor Repository stats |Metric|Value| |:-|:-| |Stars|46,813| |Forks|5.8k| |Last release|v1.6.15 (2026-10-18)| |Languages|Python 65.6%, TypeScript 33.9%, JavaScript 0.2%| |Contributors|227| |AI contributor accounts (Claude / Codex / Cursor / Copilot)|@claude (131 commits), @cursoragent (14 commits), @Copilot (6 commits)| |AI co-authored commits|Cursor: 14, Copilot: 6| Main contributor |Field|Value| |:-|:-| |Name|Bingxi Zhao| |GitHub|@pancacake| |Commits to this repo|940 (top contributor)| |Bio|Undergraduate. Always believing Researcher = Scientist + Engineer.| |Company / Location|China| |Followers|275| ────────────────────────────── 📌 How many great GitHub repos have you discovered and completely forgotten about? Don't let another useful tool disappear into your Reddit feed. Get the community's best repos, organized by category, plus key takeaways from real developer discussions, delivered to your inbox twice a week. 📬 Never lose track of great repos again. Subscribe free.

AutoSwarm: All Your Agents, Working as One (2.0k stars) What it does All text below is quoted directly from the project README. >AutoSwarm: All Your Agents, Working as One (Codex · Claude Code · DeepSeek Harness · Muse · Grok · WorkBuddy· nanobot · and more) Why AutoSwarm?: - Selects the right agents based on their capabilities; - Shares context and memory across agents; - Connects complementary tools and workflows; - Bridges data and application ecosystems; - Runs tasks in parallel for faster execution; - Coordinates agent collaboration and keeps everyone aligned; - Verifies and synthesizes results into one coherent answer. >Use AutoSwarm to read these 100 documents and summarize the key approaches and disagreements. Repository HKUDS/AutoSwarm Repository stats |Metric|Value| |:-|:-| |Stars|2,042| |Forks|370| |Last release|None| |Languages|Python 87.9%, TypeScript 9.7%, CSS 2.2%| |Contributors|9| |AI contributor accounts (Claude / Codex / Cursor / Copilot)|None| |AI co-authored commits|Claude: 40, Cursor: 1| Main contributor |Field|Value| |:-|:-| |GitHub|@LZH-YS1998| |Commits to this repo|106 (top contributor)| |Followers|87| ────────────────────────────── 📌 How many great GitHub repos have you discovered and completely forgotten about? Don't let another useful tool disappear into your Reddit feed. Get the community's best repos, organized by category, plus key takeaways from real developer discussions, delivered to your inbox twice a week. 📬 Never lose track of great repos again. Subscribe free.

جذابیتش اینه که می‌تونید به مدل‌های هوش مصنوعی مختلف وصلش کنید و برای کارهای تخصصی ازش کمک بگیرید؛ از تحلیل اسناد و تحقیق گرفته تا یادگیری و برنامه‌نویسی. البته متن‌باز بودنش به معنی رایگان بودن همه‌چیز نیست. 🔗 گیت‌هاب: HKUDS/DeepTutor

今日のGitHubトレンド HKUDS/DeepTutor 生涯にわたるパーソナライズされた個別指導を実現するDeepTutorの開発を目的としています。 個々の学習者に最適化した教育環境を提供します。