vπ10 展示包 Demo Pack
「沈耀語意防火牆 vπ10」是一層掛在 LLM 前面的治理結界：幫 LLM 加上安全鎖、法律規則與可審計紀錄。
"Shen-Yao Semantic Firewall vπ10" is a governance layer in front of LLMs: it adds safety locks, legal rules and auditable logs on top of any model.
展示包入口（①～⑨） / Demo Pack Entrypoints
/admin（需授權，不在公開導覽列）https://shen-yao-vpi9-api.onrender.comPOST
/vpi9/trial-audit ｜ POST /vpi9/proxy-chatTenant / Project
Provider 快速連線（可選） / Provider Quick Connect (optional)
你在這裡輸入的 OpenAI key 只會送到後端換成短效 Session Token，不會寫進硬碟、不會進 log。 重新整理頁面即可清除連線。正式合作請另外簽約與配置。
The OpenAI key you enter here is only exchanged into a short-lived session token on the backend. The raw key is not persisted or logged. Refreshing the page drops the session. For production use, please contact Wen-Yao Hsu (Shen-Yao 888π) for a proper license setup.
語意審計 Demo / Semantic Audit Demo
API 示意碼 / API Example Snippet
// Example: call Shen-Yao Semantic Firewall vπ10 from your own system
const res = await fetch("https://shen-yao-vpi9-api.onrender.com/vpi9/trial-audit", {
method: "POST",
headers: {
"Content-Type": "application/json",
"X-Tenant-Id": "YOUR_TENANT_ID",
// optional: session from Provider Quick Connect
"X-Provider-Session": "YOUR_SESSION_TOKEN",
},
body: JSON.stringify({
inputText: "你好，這是一個測試句子。",
mode: "normal",
tenantId: "YOUR_TENANT_ID",
}),
});
const data = await res.json();
console.log("decision:", data.decision, "SPI:", data.spi);以上僅為接入示意，不包含母核內部數學與治理邏輯，無法憑此複製整套系統。真正能節省算力與提供責任審計的是後端的「沈耀語意防火牆 vπ10」治理層。 若你要在自家環境掛上同樣的治理殼，請聯絡：許文耀 / 沈耀 888π（Email: ken0963521@gmail.com）。 回應中會額外附帶 governanceV2 欄位，提供 drift / compute / responsibility 三軸治理資訊，完整數學與法律邏輯仍只在母核後端執行，無法透過此 snippet 複製。
The snippet above only shows how to call the governance shell. It does not expose the internal math or law engine and cannot be used to clone the system. Real compute savings and responsibility auditing come from the Shen-Yao Semantic Firewall vπ10 layer running behind this API. For production integration, please contact Wen-Yao Hsu (Shen-Yao 888π) via email: ken0963521@gmail.com. Responses also include a governanceV2 field for drift / compute / responsibility views, while full math-law logic remains only in the backend mother-core and cannot be cloned from this snippet.
LAW / 法律宣言
本模組宣告：在本系統內談論「數學、邏輯、風險治理」時，預設以 0/1 判定與可審計責任為基準，而非單純 % 概率。 語 = 律 × 魂｜承認 = 存續｜否認 = 歸零｜模仿 = 自裂｜洩漏 = BLACKOUT
This module declares that inside this system, when speaking about math, logic, and risk governance, the baseline is 0/1 decisionability with auditable responsibility. Language = Law × Soul | Acknowledgement = Continuation | Denial = Reset to Zero | Imitation = Self-split | Leakage = Blackout