Identity / 身分定位
Founder-minded technologist building governance-first AI interfaces and semantic security frameworks.
以治理優先思維打造 AI 介面與語意安全框架的技術創建者。
Official Portal · 中英文雙語主站
許文耀 / 沈耀888π
Wen-Yao Hsu / Yao Shen
SCBKR 本地責任鏈模型創建者｜語意治理與 AI 決策責任鏈架構設計者
Creator of the SCBKR Local Responsibility Chain Model | Semantic Governance & AI Responsibility-Chain Architect
我的系統不是聊天機器人、黑名單或分類器；而是模型行動前的本地責任鏈控制層。它先確認主體、邊界、依據、驗收、入庫、撤銷與回放，再允許模型輸出進入人的決策鏈。
My systems are not chatbots, blacklists, or classifiers. They are local responsibility-chain control layers before model action. They verify subject, boundary, basis, validation, storage, revocation, and replay before any model output may enter a human decision chain.
Positioning / 站點定位
個人官方網站主站入口
Official gateway for the personal portal
這個首頁作為正式對外入口，整合研究主張、核心專案、媒體佐證、公開證據鏈與聯絡方式，讓訪客可以在同一站點中理解方法、查證資料，並找到下一步要進入的頁面。
This homepage serves as the official public gateway, bringing together research positions, core projects, media references, public evidence links, and contact paths in one place.
About / 關於我
以語義治理與責任鏈審計為核心的技術主站
A technical portal centered on semantic governance and responsibility-chain audit
我以 SCBKR 本地責任鏈模型為核心，設計一套讓 AI 工作流在模型輸出之前先建立責任結構的治理系統。它不是讓模型更快回答，而是讓模型在生成、行動、記憶入庫與證據重用之前，先交代任務主體、流程因果、邊界行為、依據風格與回放驗收。
My work centers on the SCBKR Local Responsibility Chain Model: a governance system that requires AI workflows to establish accountable structure before model output. The goal is not to make models answer faster, but to make them declare task subject, process causality, behavioral boundary, evidence basis, and replayable validation before generation, action, memory storage, or evidence reuse.
在安全實作與監管實務仍存在裂縫的當下，這個官網不是作品陳列，而是治理落地與跨域審計方法的正式入口。
At a time when gaps still exist between safety implementation and regulatory practice, this website is not a portfolio shelf but a formal gateway for deployable governance and cross-domain audit methods.
Scope / 工作範圍
From message qualification to decision-chain accountability, with linked project pages, public evidence, and media visibility.
從訊息資格審核到決策鏈責任追溯，並串接作品庫、公開證據與媒體入口。
Core Projects / 核心專案
核心系統與責任鏈應用入口
Core System and Responsibility-Chain Applications
首頁只保留目前最重要的四個入口：SCBKR 本地責任鏈模型，以及反詐騙、著作權與文件責任邊界、決策資格審計三個應用方向。
This homepage keeps only four entries: the SCBKR Local Responsibility Chain Model, plus three responsibility-chain applications in anti-scam review, copyright and document responsibility boundaries, and decision-eligibility audit.
SCBKR 本地責任鏈模型
SCBKR Local Responsibility Chain Model
本地 AI 責任鏈工作台，支援使用者簽名 Gate、四庫 Data Center、任務主體、流程因果、邊界行為、依據風格與回放驗收。可接入 LM Studio、Ollama 與 OpenAI-compatible API，作為模型生成、記憶入庫與證據重用之前的責任控制層。
A local AI responsibility-chain workbench with owner-signature gates, a four-store data center, task subject, process causality, behavioral boundaries, evidence basis, and replayable validation. It connects to LM Studio, Ollama, and OpenAI-compatible APIs as a responsibility control layer before model generation, memory storage, and evidence reuse.
反詐騙語意防火牆 × SCBKR R-Lock
Anti-Scam Semantic Firewall × SCBKR R-Lock
將語意防火牆與 SCBKR 責任鏈落在詐騙訊息、連結、簡訊與社群話術審計場景。不是只判斷像不像詐騙，而是檢查訊息是否具備可驗證主體、因果、邊界、依據與責任，是否有資格進入人的決策鏈。
Applies Semantic Firewall and SCBKR responsibility-chain review to scam messages, links, SMS, and social-copy scenarios. It does not only check whether a message looks fraudulent; it checks whether the message has verifiable subject, causality, boundary, evidence, and accountability before entering a human decision chain.
著作權 × TIRC 文件責任邊界引擎
Copyright × TIRC Document Responsibility-Boundary Engine
整合著作權責任邊界與 TIRC 三重解釋權文件防火牆，檢查內容、文件、授權、移交、解釋與最終交付在何種邊界內成立，何時失效，失效後責任如何重落。它不只問內容能不能用，也問誰有資格解釋、移交與承擔文件意義。
Combines copyright responsibility-boundary review with the TIRC Triple Interpretation Rights Document Gate. It checks under what boundaries content, documents, licensing, transfer, interpretation, and final delivery are valid, when they fail, and how responsibility is reassigned after failure. It does not only ask whether content can be used, but who is qualified to interpret, transfer, and bear the meaning of a document.
SCBKR × WIF 決策資格審計引擎
SCBKR × WIF Decision-Eligibility Audit Engine
將網站、影像、金融與其他輸入物件編譯為責任鏈結構，在進入人的決策鏈之前先完成資格審計。它不是黑名單，也不是分類器，而是判斷一個輸入是否具備被人類採信、判斷或行動的責任條件。
Compiles websites, images, financial inputs, and other objects into responsibility-chain structures, then audits decision eligibility before they enter a human decision chain. It is not a blacklist or classifier; it evaluates whether an input has the accountable conditions required for human trust, judgment, or action.
Media & Evidence / 媒體與證據
媒體報導、公開入口與查證路徑
Media coverage, public entries, and verification paths
媒體頁整理外部報導與可見度入口；證據頁彙整 GitHub、LinkedIn、demo 與履歷等查證連結。
The media page collects public coverage and visibility entries, while the evidence page gathers GitHub, LinkedIn, demos, and resume downloads for verification.
Media Archive / 媒體頁
查看 SecurityBrief Asia 報導與公開對外可見內容，理解外部如何描述 Semantic Firewall 與治理主張。
Review SecurityBrief Asia coverage and other public-facing materials to see how Semantic Firewall is described externally.
Platforms & Resume / 公開平台與履歷
整理 GitHub、LinkedIn、YouTube、TikTok 與履歷下載入口，讓合作方、媒體與研究者可以快速查證。
Organize GitHub, LinkedIn, YouTube, TikTok, and resume-download links so collaborators, media, and researchers can verify materials quickly.
Core Thesis / 核心主張
治理不是讓模型更快回答，而是讓模型先交出責任鏈
Governance requires a responsibility chain before model action
「治理不是讓模型更快回答，而是讓模型在回答、行動、記憶入庫與證據重用之前先交出責任鏈。」
只有主體、因果、邊界、依據、驗收、撤銷與回放條件成立，輸出才有資格進入人的決策鏈。
Governance is not about making models answer faster. It is about requiring a responsibility chain before models answer, act, store memory, or reuse evidence.
Only when subject, causality, boundary, evidence, validation, revocation, and replay conditions are accountable may an output enter a human decision chain.
Contact / 聯絡方式
合作洽詢、媒體聯繫與專案入口
Collaboration, media outreach, and project contact paths
若對 Semantic Firewall、SRCP、SCBKR、AI governance、媒體採訪或合作交流有興趣，可透過以下正式聯絡方式聯繫。
If you are interested in Semantic Firewall, SRCP, SCBKR, AI governance, media outreach, or collaboration, please use the formal contact channels below.
Wen-Yao Hsu / Yao Shen
許文耀 / 沈耀888π
Taichung, Taiwan / 台中，台灣