AI Decision System
Replayable, offline-verifiable records for AI-assisted go/no-go decisions: experiment promotions, typed model answers and human-approved actions.
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Convert anything on the internet to clean Markdown, with a searchable library, API, CLI, MCP and WebMCP.
翻譯準備中
取自儲存庫 README
正在載入完整 README。
檢視完整 README ↗The web context layer for AI agents. anymd reads public web content, normalizes it into structured Markdown, remembers what matters in a private searchable library, and makes it available anywhere through MCP, API and CLI.
The fastest way to try it: put anymd.cc/ in front of a public URL.
curl https://anymd.cc/stephango.com/saw
Connect your agent over MCP:
claude mcp add --transport http anymd https://anymd.cc/mcp
anymd.cc · Docs · API reference · MCP · Pricing
read_url, search_library, get_document…) with OAuth 2.1 or API keys, a REST API, a zero-dependency CLI, WebMCP tools in the browser, and the URL prefix: https://anymd.cc/<url> returns Markdown, JSON or an HTML preview.anymd token or *), honours domain opt-outs on every channel, and enforces per-site rate limits..md twin, plus llms.txt.# Markdown
curl https://anymd.cc/stephango.com/saw
# JSON
curl -H "Accept: application/json" https://anymd.cc/stephango.com/saw
# With an API key: counted against your credits and saved to your library
curl https://anymd.cc/api/v1/convert \
-H "Authorization: Bearer $ANYMD_API_KEY" \
-H "Content-Type: application/json" \
-d '{"url":"https://stephango.com/saw"}'
CLI:
npm i -g https://cdn.anymd.cc/cli/anymd-cli-latest.tgz
anymd https://stephango.com/saw
Claude Code:
claude mcp add --transport http anymd https://anymd.cc/mcp --header "Authorization: Bearer amd_…"
| Quickstart | First conversion to agent integration in five minutes |
| URL API | Formats, options, headers, limits |
| REST API | Every endpoint, with curl, JavaScript and Python |
| CLI | The anymd command |
| MCP · WebMCP | Agent integrations |
| Library & search | Search modes, fan-out, RRF, Jev |
| Sources | Dedicated readers and what's planned |
| API keys & roles | Scopes, presets, roles |
| Billing | Credits and plans |
| Self-hosting | Run your own copy |
The docs source is in src/content/docs/. Maintainer docs: docs/. AI coding agents: AGENTS.md. Reviewers: REVIEW.md.
anymd deploys as a single Cloudflare Worker:
OAUTH_KV, CACHE), an R2 bucket, and a 1024-dimension Vectorize index with user_id and doc_id metadata indexes.wrangler.jsonc at your account, resources and hostnames.migrations/ with wrangler d1 migrations apply.src/env.ts).npm install && npm run deploy:production.Full guide: Self-hosting. Upstream CI deploys dev to staging.anymd.cc and main to anymd.cc.
npm install
npm run dev # build + wrangler dev (staging config)
npm run typecheck
npm test
Issues and pull requests are welcome. If a page converts badly, open an issue with the URL. Read AGENTS.md for layout and conventions, and REVIEW.md for what reviews check.
MIT. Copyright (c) 2026 Digitop.ai.
Built by Digitop.ai. Contact: [email protected].