AI architecture researcher | Building intelligent agent systems & LLM infrastructure & AI Native Product Design
Repositories
alexzhu0/HippoRAG-DeepSeek-M3
https://github.com/OSU-NLP-Group/HippoRAG
alexzhu0/future-agi
Open-source, end-to-end platform for evaluating, observing, and improving LLM and AI agent applications. Tracing · Evals · Simulations · Datasets · Gateway · Guardrails. Self-hostable. Apache 2.0.
alexzhu0/AgentMo
alexzhu0/deepseek-harness
DeepSeek Harness: Everything is a Plugin.
alexzhu0/m_flow
M-Flow — Memory-augmented knowledge graph framework for AI agents
alexzhu0/skillscout-catalog-test
alexzhu0/mlx-lm
Run LLMs with MLX
alexzhu0/GOV-Report_Jina_Embeddings_v4_RAG
31government_reports RAG System
alexzhu0/sglang
SGLang is a high-performance serving framework for large language models and multimodal models.
alexzhu0/Agents-learning
alexzhu0/agent-skill-radar
Evidence-backed topic radar for choosing what Agent Skill to build next.
alexzhu0/agent-context-lab
Scan coding-agent instructions, MCP config, and repo context into a CI-friendly readiness report.
alexzhu0/agent-ready-skills
Ten practical AI-agent skills for turning messy work into clear context, evals, reviews, and launch-ready artifacts.
alexzhu0/agent-context-score
Score agent instruction context for clarity, verification quality, contradictions, and abuse-prone phrasing.
alexzhu0/mcp-tool-risk-audit
Audit MCP tool manifests for risky commands, secrets, files, network, and auth gaps.
alexzhu0/mcp-registry-diff
Compare MCP registry snapshots, normalize common formats, and detect risky server changes.
alexzhu0/metaflow
Build, Manage and Deploy AI/ML Systems
alexzhu0/keep
The open-source AIOps and alert management platform
alexzhu0/repo-to-ai-brief
Generate reviewable coding-agent briefs from repositories.
alexzhu0/promptbeat-lite
Run local-first prompt and agent regression checks.
alexzhu0/prompt-drift-watch
Detect risky prompt and coding-agent instruction drift.
alexzhu0/mcp-server-radar
Scan MCP server lists into capability and risk indexes.
alexzhu0/llm-cost-guard
Estimate prompt token cost and flag expensive LLM inputs locally.
alexzhu0/eval-case-miner
Mine reusable eval cases from AI failure logs.
alexzhu0/context-window-doctor
Find context rot in long coding-agent instruction files.
alexzhu0/agent-trace-summarizer
Turn AI agent traces into Markdown incident reports.
alexzhu0/agent-skill-lint
Lint agent skills and prompt packs for missing boundaries and risky wording.
alexzhu0/agent-runbook-kit
Generate a minimal runbook kit for AI agent repositories.
alexzhu0/hermes-agent
The agent that grows with you