Tana website demo showing a Linear issue and an email draft beside a video meeting

Tana Product Teardown: Following a Task from Notes into Meeting Work

A task written in meeting notes can keep participating in the work if it also appears on a project board, with either location editing the same content. In studying Tana, I want to start by following that task. Consider a fictional example. Project Moss needs to fix duplicate form submissions after a device has been offline. A meeting produces the task “Deduplicate submissions,” provisionally assigned to Member A. We need to know which project it belongs to, who takes it on, why it matters, and whether it actually enters a working system after the meeting. This example is for explanation and acceptance design only. It has not been executed in a real account. ...

October 23, 2026 · 16 min · 3272 words · Xinwei Xiong

Nango in Practice: Let Users Add Their Own MCP Servers in Chat

A model needs a tool to check project status. A user asks, “What’s blocking the demo project?” The model understands the question, but has no way to read the project’s state. The experience we want is straightforward: the user adds their project service, authorizes access once, then returns to the same conversation and gets an answer grounded in data. What this article verifies Sources were checked as of October 3, 2026, using Nango source and client version 0.71.12. The local examples use Node.js 22+ and have been tested for tool discovery, input validation, and request/response handling, including a local replacement for the proxy adapter. Project data is synthetic; the experiment does not call a model or execute real OAuth and proxy requests through Nango Cloud. The public-server sections provide integration steps to test in your own account. Rolling documentation, pinned source, and cloud deployments may differ; a dependency version does not prove protocol support, and customer stories establish only their stated use cases. The available evidence also does not establish that external MCP tools are automatically imported into Nango’s Action catalog. ...

October 3, 2026 · 19 min · 3888 words · Xinwei Xiong
A design canvas connected by lines to desktop and mobile interfaces

Figwright MCP: Free Figma Setup and Design Systems

Figwright is a third-party, open-source Figma MCP server. It connects an AI coding assistant to a Figma plugin through a local server, allowing the assistant to read designs and edit the canvas. It works with Figma’s free tier without a Dev Mode seat; model subscriptions or usage costs are separate. Figwright v0.7.0 documentation Set up Figwright and verify a read These connection steps were checked against Figwright v0.7.0 on November 9, 2026. The MCP server requires Node.js 20.19+ within the 20.x line, or 22.12 and later. Use matching server and Figma plugin versions. The design-system discussion and source analysis below retain their stated version boundaries. Setup and update instructions ...

September 26, 2026 · 31 min · 6526 words · Xinwei Xiong

IMStage: Synthetic Conversations for Tests and Evaluation

A chat-screenshot test needs different messages, people and timestamps. IMStage keeps those variations in editable synthetic scenes for software tests, teaching and evaluation datasets. Current project overview In IMStage , edit messages, people, avatars and device settings directly, or sign in to let the hosted Agent draft and revise a scene. Previews and exports carry an “AI-generated / Fictional” label by default. Features and usage boundaries An earlier interface from September 2026, with fictional people and synthetic photos. Current features and licensing are described below. ...

September 23, 2026 · 2 min · 311 words · Xinwei Xiong
Open Design as a four-plane design harness for coding agents

Open Design 0.16.1: A Design Harness for Coding Agents

Open Design is easy to misread. The name also belongs to the older open-design movement around shared product blueprints, but this article is about nexu-io/open-design : an open-source workspace that turns coding agents into a design production system. It is not a model, and it is not simply an image generator. It is a harness. The harness gives an agent a controlled vocabulary, reusable workflows, visual constraints, an artifact loop, and a place to inspect the result. That distinction matters because the quality ceiling still comes from the model and the operator; Open Design improves the path between intention and output. ...

July 22, 2026 · 9 min · 1791 words · Xinwei Xiong
How AI agents rebuilt a 120-post Hugo blog with a human editor

How AI Agents Rebuilt My 120-Post Blog From Scratch

What actually happened here How large a team does it take to rebuild a four-year-old blog with 120+ posts — content architecture, operations, all of it? My answer, delivered over the first half of this year: one person and a crew of agents with clearly divided jobs. “Crew” is a working metaphor, not a fixed headcount of digital employees. As of July 2026, the repository contains seven Claude skills and four workflows whose names begin with seo-; Claude Code, Codex, the Claude jobs inside GitHub Actions, and temporary review sessions are started only when a task calls for them. The human owns judgment, boundaries, and the final signature. Agents own scoped work that can be tested. The result is the site you’re reading — cubxxw.com . It isn’t merely a redesign. It turned “a blog” from a static site into a system that generates a daily report, prepares fix proposals, and exposes its own search interface. ...

July 17, 2026 · 19 min · 4026 words · Xinwei Xiong
A personal intelligence pipeline connecting monitored sources to evidence, judgment, and action

Build a Personal Intelligence System That Leads to Action

The quiet dashboard can be the dangerous one A morning briefing with no new items may mean that nothing relevant happened. It may also mean an RSS route broke, an API began returning empty pages, or a credential expired during the night. From the reader’s side, those very different states look identical: silence. That is the first lesson of a personal intelligence system. Gathering more information is easy to demonstrate; knowing whether the machine is healthy, whether its summaries are faithful, and whether a signal deserves action is the real work. The system I want is not an account that publishes automatically. It is a pipeline that can show me what changed, why it may matter, what evidence supports the claim, and which decision—if any—should follow. ...

July 15, 2026 · 12 min · 2452 words · Xinwei Xiong
The Super-Individual Stack: AI-Native Product Directions and Solo Builder Ops in 2026

The Super-Individual Stack: AI-Native Product Directions and Solo Builder Ops in 2026

“Software is eating the world.” — Marc Andreessen, 2011 “Now AI is eating software—and the question for the rest of us is: what’s left for one human, alone, in front of a screen?” — me, asking myself one night in 2026. Prologue: How Big Does One Person Need to Be? In February 2026, I ran my first complete overnight agent. I set a prompt, dropped it into Claude Code in a loop, and went to sleep. At 7 a.m. the next morning, what I saw on the screen was: 6 commits, 4 PRs, 3 auto-rolled-back failures, and a research brief I hadn’t even read myself. ...

June 24, 2026 · 21 min · 4314 words · Xinwei Xiong
A wide schematic of context engineering: the Write / Select / Compress / Isolate pillars feeding an AI, a laptop with notes, and a local-first world line

Context Engineering: The New Foundation for AI Agents

Context engineering is the set of strategies for curating, ordering, and evicting the tokens that enter an LLM’s context window on each inference call—system instructions, retrieved documents, conversation history, tool definitions, and memory. The short version is this: prompt engineering optimizes the wording of an instruction; context engineering optimizes the wiring of the whole window. Anthropic has articulated this distinction in its engineering guidance, and Karpathy helped popularize the term. The rest of this article takes the emerging discipline apart. ...

June 22, 2026 · 18 min · 3655 words · Xinwei Xiong
A technical diagram with a tiny agent loop at the center, surrounded by concentric rings of the eight pillars: orchestration, context, memory, tools, reliability, evaluation, cost, governance

Agent Engineering Harness: The Eight Pillars Behind the 98.4%

A small model loop becomes a system only after we decide what it may remember, touch, spend, and survive. An agent demo can be a loop around an API call. A production agent is the surrounding discipline: state, tools, permissions, recovery, evaluation, and cost. This article offers a map of that discipline rather than a framework tutorial. Each pillar is described through the gap it fills, a minimal implementation, and the boundary where it stops helping. ...

June 17, 2026 · 20 min · 4223 words · Xinwei Xiong
GPT Researcher pipeline from query planning and retrieval to a cited report

GPT Researcher Guide: Python, Docker, MCP, Costs & Limits

A long report can look like certainty while merely arranging uncertainty more elegantly. The useful question is not how many pages an agent writes, but how a claim entered the report and whether a reader can walk back to its source. The short verdict GPT Researcher is a good fit when a team needs a programmable research pipeline, source traceability, and deployment control. It is not a truth machine. It automates planning, retrieval, context assembly, and report writing; it does not make weak pages authoritative or make every citation support the sentence beside it. ...

April 14, 2025 · 10 min · 2024 words · Xinwei Xiong