Jask MCP Hub

One Model Context Protocol endpoint serving every Jask / UZEN Labs product's content. Read-only, no auth.

https://mcp.jask.dev/mcp

Sources

site id	site	content	items	map
`onchaindiary`	theonchaindiary.com	Web3 on-chain security education (articles + glossary, EN/ZH)	419	.md
`zensink`	zens.ink	ZensInk SEO workflow docs + journal (EN/ZH/JA/ES)	126	.md
`blog`	blog.jask.dev	Jask's personal tech blog (ZH)	30	.md
`liteink`	liteink.co	Astro template shop — premium/free themes, docs, blog (EN)	38	.md
`revolutio`	revolutio.si	AI disruption intelligence — industry analyses, blog only (EN)	18	.md
`uzenlabs`	uzenlabs.com	UZEN Labs build log + product pages (EN)	25	.md
`urgenth`	www.urgenth.com	Customs & HS classification intelligence — guides + HS Code Finder (EN)	20	.md
`docsuzen`	docs.uzenlabs.com	UZEN Labs platform docs — principles, product cards, MCP & AI-agent guides (EN)	9	.md

Tools

list_sites() — overview of connected sources
search(query, site?, lang?) — cross-site keyword search (EN/ZH)
read(site, id) — full text of one item as markdown

Try it

curl https://mcp.jask.dev/mcp \
  -H 'content-type: application/json' \
  -d '{"jsonrpc":"2.0","id":1,"method":"initialize",
       "params":{"protocolVersion":"2025-03-26","capabilities":{},"clientInfo":{"name":"curl","version":"0"}}}'

# then search + read
curl https://mcp.jask.dev/mcp -H 'content-type: application/json' \
  -d '{"jsonrpc":"2.0","id":2,"method":"tools/call",
       "params":{"name":"search","arguments":{"query":"wallet drainer"}}}'

Use with an MCP client

{
  "mcpServers": {
    "jask": {
      "type": "http",
      "url": "https://mcp.jask.dev/mcp"
    }
  }
}

No setup? Paste this into any AI chat

Copy the prompt below and send it to any AI assistant — it will set things up or tell you exactly what to do.

setup prompt

Help me connect to an MCP server to search and read a collection of technical content in English and Chinese.

Server endpoint (Streamable HTTP, no auth required, connect directly):
https://mcp.jask.dev/mcp

It exposes 3 read-only tools:
- list_sites() — list all connected content sources with item counts
- search(query, site?) — search across all sites; query works in English and Chinese; optional site limits to one source
- read(site, id) — fetch the full text of one item; site and id come from search results

Content sources (~590 items, auto-updated):
- theonchaindiary.com — Web3 on-chain security education, articles + glossary (EN/ZH)
- zens.ink — SEO workflow docs (EN/ZH/JA/ES)
- blog.jask.dev — indie dev tech blog (ZH)
- liteink.co — Astro template shop, themes + docs (EN)
- revolutio.si — AI disruption intelligence, industry analyses (EN)
- uzenlabs.com — product build logs (EN)
- urgenth.com — customs & HS classification guides + HS Code Finder (EN)
- docs.uzenlabs.com — UZEN Labs platform docs, principles + MCP/AI-agent guides (EN)

If you support MCP: add the server above now, call list_sites to confirm the connection, then tell me what I can search.
If you do not support MCP: tell me which client app I should use to connect to this server, with concrete steps.