Generated Computers
Locals: an Apple Silicon engine that runs local LLMs up to 3x faster. A product of Generated Computers.
翻譯準備中
尚未偵測到更新。GitHub 發布版本後會顯示在這裡。
個人 GitHub 帳號下公開儲存庫的擁有者可以登入並驗證所有權,自行管理介紹,無需管理員核准。
登入並產生驗證檔案,在 30 分鐘內將其提交至此儲存庫的預設分支。其他分支或分叉儲存庫中的檔案無效。然後在下方檢查檔案,即可無需管理員核准直接管理專案。請保留檔案以維持管理權限。
如需登入,將前往登入頁面。登入後請回到此處,再次點擊所選的驗證按鈕。
在專案資料夾中執行 /nomorevibe verify,即可驗證所有權並自行管理介紹。
一分鐘完成驗證Requstory CLI: map how a company works from its website, find the gaps, check ease scores and AI-agent readiness.
翻譯準備中
取自儲存庫 README
正在載入完整 README。
檢視完整 README ↗Requstory from the command line. Map how a company works from its public website (customer journey, teams, processes, handoffs), find the gaps, ask questions, and check public ease scores and AI-agent readiness.
It also runs order to cash for a B2B online store with four AI agents (process orders, ship orders, invoice and collect payments, handle returns) that stop and ask you about credit limits, short stock, packing, reminders, unclear payments and returns.
https://requstory.com/mcpnpm install -g requstory
# or run once
npx requstory --help
Node 20 or later. No dependencies.
# Public, no key needed
requstory companies posthog
requstory score posthog-com
requstory readiness example.com
# Your maps: create a key in your Requstory profile (API keys)
requstory login rqs_your_key
requstory map example.com
requstory maps
requstory show <mapId>
requstory ask <mapId> "Who owns billing?"
requstory ask <mapId> "What should we fix first?" --ai # Requ AI, uses credits
Add --json to any command for the raw reply. The key can also come from REQUSTORY_API_KEY. login saves it to ~/.config/requstory/config.json (readable only by you).
| Command | What it does | Key |
|---|---|---|
companies <query> | Search mapped and scored companies | No |
score <slug> | Ease score (0–100), ranking place and gaps | No |
readiness <website> | Can AI agents read and use a website? | No |
map <website> | Map a website and save it to your account (5 a day) | Yes |
job <jobId> | Status of a map still running | Yes |
maps | Your saved maps | Yes |
show <mapId> [--full] | A map in brief, or the whole model | Yes |
ask <mapId> <question> [--ai] | Answers with sources | Yes |
account | Plan, credits and maps left today | Yes |
o2c demo | Start the made-up store's month | No |
o2c start <business.json> | Start a day of your own business | Yes |
o2c run / exceptions / decide / order / export | Drive a run and get the record | Demo: no |
# A month of a made-up store, no key needed
requstory o2c demo
requstory o2c run <runId> # agents work until they need you
requstory o2c exceptions <runId> # what they are asking
requstory o2c decide <runId> <exceptionId> accept
requstory o2c run <runId>
requstory o2c export <runId> --out=results
# A day of your own business (key; $1.00 an order from your credits, model cost included)
requstory o2c start business.json --map=<mapId>
business.json holds what the run page's folder holds. Every list is optional except customers:
{
"company": "Bay Glassworks", "symbol": "£", "vat": "20%", "termsDays": 30, "nextInvoiceNo": 501,
"customers": [{ "id": "C01", "name": "Garcia Inc", "email": "[email protected]", "creditLimit": 5000, "openBalance": 0 }],
"products": [{ "sku": "GL-12", "name": "Pebbled goblet 12 oz", "price": 8.99, "stock": 40 }],
"orders": [{ "file": "orders/po-7781.jpg" }, { "name": "email from Garcia", "text": "Please send 12 x Pebbled goblet 12 oz. PO 7782" }],
"openInvoices": [{ "no": "INV-77", "customer": "Garcia Inc", "total": 240, "paid": 0, "due": "2026-10-20", "date": "2026-10-06" }],
"bank": [{ "from": "GARCIA INC", "ref": "INV-77", "amount": 240 }],
"returns": [{ "message": "The glasses on invoice INV-77 arrived broken." }],
"returnsPolicy": "Damaged in transit: full refund or credit within 30 days.",
"instructions": { "finance": "Sign emails \"Accounts team\"." },
"approvals": { "ordersOver": 2000, "overCreditLimitUpTo": 100, "refundsUpTo": 50, "returnsWithinDays": 30, "shortPaymentWriteOffUpTo": 2 }
}
Order scans (file) are JPEG or PNG under 3 MB, relative to the JSON file. Without approvals the agents ask you about every credit overrun and refund; with them, orders above ordersOver always come to you, the code (not the model) lets through overruns up to overCreditLimitUpTo and refunds up to refundsUpTo whose amount it has worked out itself, returns for invoices older than returnsWithinDays (or with no date) always come to you, and a payment that leaves shortPaymentWriteOffUpTo or less owed closes the invoice. 0 means no rule. Each order is charged once, the first time it is invoiced, collected or has a return settled; a $10 credit pack covers 10 orders, and a run stops before an order your credits cannot pay for. Sums, dates, stock and credit are plain code. Nothing is sent: emails are drafts in the export. Runs are deleted 2 days after the last change.
import { createClient } from "requstory";
const requstory = createClient({ key: process.env.REQUSTORY_API_KEY });
const { companies } = await requstory.companies("stripe");
10 calls a minute without a key, 60 with one. 5 new maps a day. 3 AI-readiness checks an hour.
MIT © 3D7 Technologies Ltd