Skip to content
View however-yir's full-sized avatar
🛠️
Building practical AI products · Open to opportunities
🛠️
Building practical AI products · Open to opportunities

Block or report however-yir

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
however-yir/README.md

Hi, I'm however-yir

AI Engineer & Java Backend Developer working on knowledge retrieval, business agents, and backend reliability.

专注 AI 应用工程与 Java 后端:让知识问答有据可查,让业务 Agent 的动作可控,让系统改动有验证依据。

Featured Projects

Project Engineering focus Explore
KnowledgeOps Agent Spring AI platform prototype: tenant-scoped RAG, asynchronous document ingestion, answer citations and evaluation. Demo · Design · Evidence
Tianji AI Agent Course-advisor Agent: intent routing, governed order previews, execution budgets and SSE contracts on an existing business backend. Demo · Design · Evidence
NebulaKB Django knowledge operations: document ingestion, retrieval feedback and low-quality answer review. Demo · Design · Evidence
ForgePilot Studio OpenHands-based workbench: task console, audit tooling and experimental governance modules. Demo · Design · Evidence
Microservices Lab Online Boutique extension: AI shopping assistant, local model fallback and Kubernetes integration. Demo · Design · Evidence
ragproof Framework-agnostic RAG evaluation CLI: recall@k, MRR, faithfulness, citation metrics, regression gates and CI integration. Demo · Design · Evidence

Start with KnowledgeOps for the Java/RAG platform, then Tianji for business Agent execution.

For inherited code and project-specific changes, see the project scope and provenance guide. ForgePilot's core governance modules are experimental and are not wired into the default OpenHands execution path; its design links describe that boundary.

Reading routes · How to interpret the evidence

Open Source Contributions

10 merged external pull requests across Dify (4), Spring AI Alibaba examples (4), LangChain4j (1), and MiniMax CLI (1), covering MCP tool schema handling, framework backend refactoring, documentation, and configuration templates.

  • LangChain4j: #6320 added merging of allOf object sub-schemas when converting MCP tool input schemas, with regression tests.
  • Dify: #41882 passed database session into ApiToolProvider.user; #41883 removed legacy db.session wrappers on Message feedback and annotation accessors; #41885 passed database session into AppAnnotationSetting.collection_binding_detail; #41886 removed legacy db.session wrappers on additional Message accessors.
  • Spring AI Alibaba examples: #452 quick-start matrix; #453 MCP/RAG/tool-calling environment templates; #457 onboarding and troubleshooting; #458 additional DashScope environment templates.
  • MiniMax CLI: #84 corrected the video download example and clarified credential locations.

Contact

  • Email: liuhowever@gmail.com
  • Interests: AI products, backend engineering, consumer tech, history, travel, coffee shops, food, films, and football

Pinned Loading

  1. knowledgeops-agent knowledgeops-agent Public

    Production-oriented Spring AI platform prototype for RAG, tool calling, async ingestion, JWT/RBAC, and observability.

    Java 191 17

  2. tianji-ai-agent tianji-ai-agent Public

    Production-oriented multi-agent course advisor with deterministic action governance, offline regression evaluation, SSE contracts and observable tool execution.

    Java 45 4

  3. forgepilot-studio forgepilot-studio Public

    OpenHands-based AI engineering workbench with experimental control-plane, audit, multi-model, and deployment tooling.

    Python 22 1

  4. nebula-kb nebula-kb Public

    Local-first knowledge asset lifecycle platform — ingestion, governance, retrieval, feedback loops & ops dashboard. Built with Django, PostgreSQL (pgvector), Redis.

    Python 3 1

  5. however-microservices-lab however-microservices-lab Public

    Multi-language microservices lab with AI integration, 12 services on Kubernetes with gRPC and HTTP.

    Python 3

  6. ragproof ragproof Public

    Framework-agnostic RAG evaluation & regression-testing CLI — recall@k, MRR, faithfulness, citation metrics, CI gate

    Python 2