Knowledge Share
Technical articles, tutorials, and insights
Enterprise Knowledge Base (02): Classic Vector RAG Benchmark — QAnything vs LightRAG
Same 89 questions, two frameworks: QAnything v2 and LightRAG 1.5.6. This article covers the full journey from deployment to results — including Docker GPU mounting, Milvus crash recovery, user_id isolation pitfalls, and the actual numbers across single-hop, multi-hop, and boundary refusal.
Open Source Project #190: claude-video — Give Claude Eyes to Watch Videos, One Command to Analyze YouTube/Loom/Local Files, Subtitle-First Zero-Download, Scene-Aware Frame Dedup, Auto Token Budget
Brad Bonanno's Claude Code skill plugin that gives Claude the ability to analyze any video via the /watch command. Seven-step pipeline: yt-dlp download → ffmpeg frame extraction → Whisper transcription → Claude context injection. Subtitle-first optimization (use existing subtitles, skip download entirely), 16×16 thumbnail MAD algorithm for frame deduplication, automatic frame budget scaling by video duration. Four detail modes: transcript/efficient/balanced/token-burner. Works with Claude Code, Codex, Cursor, Copilot, Gemini CLI, and 50+ other agent hosts. 15.2k Stars, MIT.
Open Source Project #191: gstack — YC CEO Garry Tan's Open-Source AI Engineering Team, 23 Expert Role Commands Covering the Full Sprint from Idea to Production
Y Combinator CEO Garry Tan's Claude Code skill set that turns Claude Code into a staffed engineering team. 23 professional role commands covering planning, design, review, testing, and release: /office-hours (YC-style product challenge), /plan-ceo-review, /design-shotgun (4-6 design variants for comparison), /qa (real Playwright browser testing), /cso (OWASP + STRIDE security audit), /ship (sync→test→audit→push→PR). Skills pass context to each other, supports 10-15 concurrent sprints, works across Claude Code/Codex/Cursor and 10 other agents. 128k Stars, MIT.
Open Source Project #187: Pi — Philosophy-Driven Minimal AI Coding Agent, 86k Stars, 30+ LLM Providers, Unlimited Extensibility
Minimal philosophy-driven AI coding agent harness. 5 npm packages: unified LLM API (30+ providers), agent runtime, TUI library, coding agent CLI, telemetry contracts. Four modes (interactive/print/JSON/RPC/SDK), session branching tree, TypeScript extension system, Skills/Prompt Templates/Pi Packages ecosystem. Deliberately excludes MCP, sub-agents, permission popups, plan mode. TypeScript, MIT, 86k Stars.
Open Source Project #188: AI Agents in Depth — Li Bojie's Complete Open-Source AI Agent Book, 10 Chapters, 95 Experiments, 35k Stars
Li Bojie's fully open-source book 'AI Agents in Depth: Design Principles and Engineering Practice.' Core formula: Agent = LLM + Context + Tools. 10 chapters building progressively: context engineering, user memory and knowledge bases, MCP tool protocol, coding agents, evaluation frameworks, model post-training (SFT/RL), continuous evolution, multimodal interaction, multi-agent collaboration. 95 hands-on experiments, 13 language versions, free PDF/EPUB download. Python, Apache-2.0, 35k Stars.
Open Source Project #189: DeepTutor — Agent-Native Lifelong Learning Workspace, 3-Layer Memory + Multi-Engine RAG + Partners, 33k Stars
Agent-native AI learning workspace from HKUDS (HKU Data Intelligence Lab). Core: a single agent loop drives all modes (Chat/Quiz/Research/Visualize/Solve/Mastery Path). Three-layer memory (L1 event traces / L2 surface summaries / L3 cross-surface synthesis) with a visual Memory Graph tracing every claim to its source. Multi-engine knowledge bases: LlamaIndex/PageIndex/GraphRAG/LightRAG/Obsidian. Partners: persistent AI companions with IM channels (Slack/Discord/Telegram/WeChat/Feishu, etc.). My Agents: drive local Claude Code/Codex as subagents. Python + Next.js, Apache-2.0, 33k Stars.
Open Source Project #186: Open Ontologies — Rust MCP Server for AI-Native Ontology Engineering, 70+ Tools, Single Binary, No JVM, Builds in 5 Minutes What Takes 4 Hours in Protégé
The Tesseract Academy's Rust MCP server that turns ontology engineering into AI-agent-callable capabilities. 70+ tools covering validation, reasoning, alignment, lifecycle governance, and RAG augmentation. Three-layer architecture: Dynamics (atomic ops + OWL-RL), Causal (causal identification), Planner (PDDL compilation). Tauri 2 + React 19 Studio desktop app with AI chat panel for building production OWL ontologies. Pizza ontology benchmark: 96% coverage in 5 minutes vs. 4 hours manual in Protégé. OAEI anatomy alignment precision 0.960 (3rd globally). Built-in marketplace of 33 standard ontologies including full IES support for UK National Digital Twin Programme. 341 Stars, MIT.
Enterprise Knowledge Base (00): Build the Evaluation Dataset Before Writing a Single Line of Code
Comparing six open-source RAG frameworks means nothing if each is tested on different documents with different questions. This is the zero article in the series — building a unified test set first, so every subsequent benchmark runs on the same 89 questions. The article covers the decision path from 'just use BEIR' to 'synthesize domain-specific questions with an LLM,' plus the complete code for generating the evaluation set.
Enterprise Knowledge Base (01): Why RAG Is Just the Starting Point
The real challenge in enterprise knowledge bases isn't technology selection — it's bad data quality, fragmented knowledge, mixed modalities, and knowledge decay. These are problems RAG can't solve. This article maps the complete technology landscape: from classic vector RAG to graph RAG, hypergraph RAG, and Agent-native knowledge systems, with what each generation actually fixes.
Open Source Project #174: AirLLM — Run 70B Models on 4GB GPU, 405B on 8GB, and 2.8-Trillion-Parameter Kimi K3 on 3.7GB
AirLLM breaks the VRAM wall with layer-wise inference: splits transformer models into per-layer shards on disk, loads one layer at a time during inference, releases memory immediately after. No quantization, distillation, or pruning required. Runs 70B to 2.8T parameter models on consumer GPUs. Supports Llama 4, Qwen3, DeepSeek-V3/R1, Kimi K3. Optional 4bit/8bit block-wise compression for ~3x speedup. 25.8k Stars, Apache 2.0, pip install airllm.
Open Source Project #175: Buzz — Block Rebuilds Team Collaboration on Nostr, Where AI Agents Hold Their Own Cryptographic Identity
Block's open-source team workspace launched July 21, 2026, where humans and AI agents collaborate as equal members in shared channels. Built on Nostr: every agent holds its own cryptographic keypair — identity and history are portable across any Nostr-compatible system. Rust backend, Tauri desktop, supports Claude Code, Codex, Goose. YAML workflow engine, buzz-cli (JSON in/out, machine-first design for LLM tool calls), NIP-34 Git integration. 21.3k Stars, Apache 2.0, self-hostable.
Open Source Project #176: Better Harness — A Five-Dimension Workflow Evaluator for AI Coding Agents That Reviews the Loop, Not the Diff
QoderAI's open-source tool that converts project and session evidence into prioritized workflow improvements for AI coding agents. Evaluates the workflow around the agent — not final code quality — across five dimensions: task understanding, controlled execution, change validation, reliable delivery, and learning capture. Three independent evidence agents analyze in parallel; a lead agent synthesizes findings into HTML/Markdown/JSON reports. Supports Claude Code, Codex, GitHub Copilot, Cursor, Qwen Code. 1.5k Stars, MIT license.