Xinwei Xiong (cubxxw) · An Open Archive of Practice and Thought

An AI content creation workflow for turning knowledge cards into audience-ready work, with evidence gates, human review, feedback, and measurable outcomes.

An AI knowledge base workflow that turns notes into verified, reusable capability through evidence gates, retrieval, review, and reversible retirement.

An AI note-taking workflow for turning text, voice, screenshots, code, and decisions into traceable records that can be reviewed and promoted to knowledge.

An AI information filtering workflow for capturing useful sources without building a noise archive, with privacy gates, verification, and human review.

A practical AI knowledge workflow for moving information through traceable records and verified knowledge into audience-ready creation with human review.

Measure GEO without proxy myths: combine Google generative AI impressions, repeated citation audits, referral analytics, and conversions in one clear protocol.

A real-data GEO case study using Search Console and Lighthouse to separate noisy impressions from useful demand, protect a domain move, and plan a Hugo rebuild.

A practical 2026 guide to off-site GEO trust: author identity, third-party evidence, ethical community distribution, and repeatable citation measurement.

A practical guide to clearer technical articles: lead with useful answers, use honest evidence, add valid schema, build internal links, and test AI visibility.

A practical 2026 guide to how Google, Perplexity, and ChatGPT retrieve and cite sources, separating official facts from inference and showing how to test GEO.

An evidence-led guide to generative engine optimization in 2026: what Google, OpenAI, Perplexity, Pew, Bain, and the original GEO paper actually support.

In June I came back from Laos to Shenzhen.

A commit-pinned audit of open-lovable’s text protocol, Agentic Search, Morph edit path, and E2B/Vercel sandbox boundaries for practical AI app builders.

Anxious attachment is not about loving too much.

In the Chinese context, what people often call avoidant personality is more precisely avoidant attachment.

Quality and time in companionship are not a choice between two things.

As the model layer commoditizes and 40% of enterprise AI projects are headed for cancellation, the super-individual finally has the full primitive set to cut into a $4.6T services market.

A source-based audit of Relay’s coordinator, five domain agents, LangGraph HITL, browser delivery, durability gaps, and remaining 2026 migration debt.

Five lenses—psychology, biology, anthropology, history, sociology—dissect love, and then Buddhism's dependent origination gathers all five into a single circle.

Learn how context engineering improves AI agents through retrieval, compaction, memory, and token budgeting—and why high-signal context beats a larger window.

Agent Engineering Harness explained through eight production pillars: orchestration, context, memory, tools, reliability, evaluation, cost, and governance.

In May I did not write a single flomo note.

April 2026 Thought Notes - Observations from living in Laos during Songkran, engineering philosophy in the AI era, and explorations of recording, understanding, and trust.

A practical framework for continuous AI agent identity, covering memory, permissions, provenance, multi-agent handoffs, and cross-session evaluation design.

When AI can replace our thinking, execution, and even creation, where does the weight of "self" come from?

March 2026 thought notes, 468 entries across 31 days.

Reflections from a stay in Lhasa, February to March 2026.

3,614 notes, fourteen months, one person's attempt to live authentically.

When AI can replace our thinking and execution, where is "me"?

February 2026 deep thought notes with 365 entries covering travel experiences (Chiang Mai, Pokhara), ecological protection, urban governance observation, AI product thinking, Buddhist philosophy, and self-discovery.

January 2026 deep thought notes with 388 entries covering AI product design, finance, biology, Buddhism, social observation, and self-discovery.

December 2025 deep thought notes with 380 entries documenting pilgrimage experiences, Zen philosophy, and self-discovery journeys.

Summer and autumn 2025, two trips to Japan.

November 2025 deep thought notes, documenting reflections across multiple dimensions: psychological mechanism exploration, urban cultural observation, historical philosophy contemplation, AI product design, religious philosophy understanding, social pheno…

October 2025 deep thinking notes, recording multi-dimensional reflections on urban development, tech product thinking, personal cognitive growth, and social phenomena.

September 2025 Thought Notes - Exploring the intersection of technology and philosophy during the early autumn harvest season, recording deep reflections on the AI era, personal growth, cognitive upgrades, and the wisdom of maintaining inner clarity in a…

August 2025 monthly reflection - harnessing late summer energy, building momentum for autumn goals, and reflecting on the transformative power of consistency.

July 2025 monthly reflection - diving deep into summer adventures, philosophical contemplations, and the art of balancing action with reflection.

June 2025 monthly reflection - embracing the summer solstice energy, strategic planning for the second half of the year, and celebrating progress made.

A first-person account of observing automatic thoughts, emotions, and self-narratives—and using metacognitive journaling to make more deliberate choices without turning self-awareness into avoidance.

May 2025 monthly reflection - exploring mid-year insights, creative breakthroughs, and the evolving landscape of personal and professional growth in the AI era.

A practical 2026 guide to Mem0 OSS v3: ADD-only extraction, hybrid retrieval, entity linking, governance, evaluation, and when not to use it in production.

A practical analysis of Microsoft UFO³, tracing its path from Windows UI automation to Desktop AgentOS and cross-device orchestration, with limits and guidance.

Build a reliable Argo CD delivery model: understand Applications, ApplicationSets, sync and rollback semantics, version pinning, RBAC, and production security.

April 2025 monthly reflection - capturing the essence of spring growth, new beginnings, and the continuous journey of learning and self-improvement.

A practical guide to LLM recommendation systems: semantic retrieval, ranking, RAG, exploration, evaluation, feedback loops, and production trade-offs.

A practical 2026 guide to Gemini Notebook: grounded research, citations, Studio outputs, plan limits, privacy, mobile tradeoffs, and reliable workflows.

A 2026 guide to test-driven development for AI and LLM systems, covering deterministic tests, evals, calibrated graders, CI risk gates, and monitoring.

Learn MarkItDown 0.1.6 for document-to-Markdown workflows, choose local, OCR, or Azure conversion, and test output quality before building a RAG pipeline.

A practical 2026 guide to LangGraph architecture: choose the right agent layer, build checkpointed StateGraphs, and recover safely from production failure.