Building a few AI products that turn complex judgment into traceable pipelines: mbabrand (a jury of agents scoring brands), bossagent (a decision agent for leaders), and oaf.world (an AI investment-research workbench). I keep 30-some Claude Skills on my machine, and post notes on my WeChat account “Zhang Lu's Murmurs.”https://zhanglu.net/https://zhanglu.net/en/posts/openworker-zhhttps://zhanglu.net/en/posts/openworker-zhThe hard part wasn't the strings — before 0.1.7 one auto-update turned it back into English. Plus two designs worth stealing from OpenWorker's source: concurrent reads with sequential writes, and four permission levels.Tue, 04 Aug 2026 00:00:00 GMTagentlocalisationsource analysisdesktophttps://zhanglu.net/en/posts/free-model-porthttps://zhanglu.net/en/posts/free-model-portA list records claims, not evidence. FreeModel Port splits candidates from runnable models, with one real request drawing the line — plus why a unified router is inherently a high-privilege centre.Tue, 04 Aug 2026 00:00:00 GMTLLMarchitecturelocal-firsttrade-offshttps://oaf.asia/stories/free-model-port/https://oaf.asia/stories/free-model-port/A feature story on FreeModel Port: continuously discovering free LLMs, verifying them with real requests, and connecting them safely to Codex and AI CLIs through a local endpoint.Sun, 02 Aug 2026 00:00:00 GMTLLMAI CLIfree modelslocal-firsthttps://zhanglu.net/en/posts/ai-interviewhttps://zhanglu.net/en/posts/ai-interviewAchievement reviews fail two ways: the smooth talker wins, and plans get counted as results. This method removes both with a 1–2 minute voice interview, adaptive follow-ups and one scoring ruler.Sat, 01 Aug 2026 00:00:00 GMTAI reviewvoiceachievement scoringmethodhttps://zhanglu.net/en/posts/brain-radarhttps://zhanglu.net/en/posts/brain-radarA brain-science × AI opportunity radar for the class of 2027. The scarce thing is not a list of opportunities but the ordering — what you have to do first for the next thing to be possible.Sat, 01 Aug 2026 00:00:00 GMTbrain sciencejob huntingaggregationradarhttps://zhanglu.net/en/posts/excel-ai-analysthttps://zhanglu.net/en/posts/excel-ai-analystThe first obstacle to putting a manual Excel process on AI rails is not compute, it is trust — you will not upload your real cost sheet to a website. So step 0 is pure front-end and works offline.Sat, 01 Aug 2026 00:00:00 GMTExcelreverse engineeringin-browserdata safetyhttps://zhanglu.net/en/posts/siliconforgehttps://zhanglu.net/en/posts/siliconforgeMost "AI writes your code" products assume the user is a developer. SiliconForge drops the entry cost to nine multiple-choice questions, and pays for it by splitting "the AI says it is done" from "the tests actually ran".Sat, 01 Aug 2026 00:00:00 GMTAI-nativesoftware factoryCloudflarePRDhttps://zhanglu.net/en/weekly/2026-w29https://zhanglu.net/en/weekly/2026-w29Three decision-agent products each shipped a major release (MBA v0.6 / Boss v1.18 / OAF v0.43), and the personal site established its main narrative, "designed for the AI-native organization." A redacted, public weekly.Sun, 19 Jul 2026 00:00:00 GMTweeklyagentdecisionhttps://zhanglu.net/en/posts/c-suite-designhttps://zhanglu.net/en/posts/c-suite-designBoss / MBA Brand / OAF look unrelated, but they share one skeleton: research → deliberate → score → attribute. Here's that skeleton, and why an AI-native organization's C-suite needs it.Sat, 18 Jul 2026 00:00:00 GMTc-suiteagentdecisionclaudehttps://qcc-agent.pages.dev/articlehttps://qcc-agent.pages.dev/articleWrapped Qcc's 146 APIs into 6 MCP servers plus 8 due-diligence Skills, so the Agent can call them directly without handling signing, pagination, or field semantics itself.Wed, 10 Jun 2026 00:00:00 GMTmcpagentqccclaudehttps://zhanglu.net/en/posts/agent-clihttps://zhanglu.net/en/posts/agent-cliAdded 9 JSON endpoints plus a zero-dependency CLI to zhanglu.net, so an agent can just run npx zhanglu-net list skills. No MCP, no SDK, shipped in 1.5 hours.Tue, 09 Jun 2026 00:00:00 GMTskillagentclaudeinfrahttps://qiji-roadshow-2026.pages.dev/storyhttps://qiji-roadshow-2026.pages.dev/storyUsed /demo-day-dossier plus 7 parallel research agents to turn all 56 projects from the MiraclePlus 2026 roadshow into a publishable, investment-grade DD overview page in one night.Mon, 08 Jun 2026 00:00:00 GMTskillclaudehttps://zhanglu.net/en/projects/ai-interviewhttps://zhanglu.net/en/projects/ai-interviewTurn one short conversation into a story your work can tellThu, 01 Jan 2026 00:00:00 GMThttps://zhanglu.net/en/projects/free-model-porthttps://zhanglu.net/en/projects/free-model-portA discovery, live-verification, and local access layer for free LLMs worldwide, built to connect Codex and other AI CLIs safelyThu, 01 Jan 2026 00:00:00 GMThttps://zhanglu.net/en/projects/aiphttps://zhanglu.net/en/projects/aipA compounding personal Wiki knowledge base continuously written and maintained by an LLM—knowledge is compiled once, then kept fresh forever (a Wiki is not RAG)Thu, 01 Jan 2026 00:00:00 GMThttps://zhanglu.net/en/projects/bosshttps://zhanglu.net/en/projects/bossA strategic OS for the CEO—hand your important judgments to a multi-juror deliberation that is citable, scorable, and falsifiable, with 30/90/365-day attribution checksThu, 01 Jan 2026 00:00:00 GMThttps://zhanglu.net/en/projects/brain-radarhttps://zhanglu.net/en/projects/brain-radarFrom experimental psychology toward brain science × AIThu, 01 Jan 2026 00:00:00 GMThttps://zhanglu.net/en/projects/oafhttps://zhanglu.net/en/projects/oafA satellite-internet decision agent + an AI investment-research workbench for the US / A-share / HK markets—numbers come from tools, narrative from the model, gaps left blank not fabricatedThu, 01 Jan 2026 00:00:00 GMThttps://zhanglu.net/en/projects/excel-ai-analysthttps://zhanglu.net/en/projects/excel-ai-analystUpload a spreadsheet, see through its structure firstThu, 01 Jan 2026 00:00:00 GMThttps://zhanglu.net/en/projects/qcc-agenthttps://zhanglu.net/en/projects/qcc-agentDrop Qichacha into any AI agentThu, 01 Jan 2026 00:00:00 GMThttps://zhanglu.net/en/projects/openworker-zhhttps://zhanglu.net/en/projects/openworker-zhHand the AI the outcome, not just the answerThu, 01 Jan 2026 00:00:00 GMThttps://zhanglu.net/en/projects/mbabrandhttps://zhanglu.net/en/projects/mbabrandTurn brand influence into a monitorable / scorable / reviewable intelligent asset—43-juror deliberation with sentiment-driven versioned re-auditsThu, 01 Jan 2026 00:00:00 GMThttps://zhanglu.net/en/projects/qiji-roadshow-2026https://zhanglu.net/en/projects/qiji-roadshow-202656 roadshow projects, from on-site photos to a publishable DD panorama page, all through one pipelineThu, 01 Jan 2026 00:00:00 GMThttps://zhanglu.net/en/projects/shanghaihttps://zhanglu.net/en/projects/shanghaiAn unofficial recap site turning the 30 talks from the 2026 AI Vision Forum (Shanghai) into minutes, a knowledge base, and a knowledge graphThu, 01 Jan 2026 00:00:00 GMThttps://zhanglu.net/en/projects/siliconforgehttps://zhanglu.net/en/projects/siliconforgeRequirements in, software outThu, 01 Jan 2026 00:00:00 GMThttps://zhanglu.net/en/projects/tui3https://zhanglu.net/en/projects/tui3Several Chinese personal websites I was part of between 1999 and 2007, rebuilt in full from server archives into a static site, preserved foreverMon, 01 Jan 2007 00:00:00 GMT