Tvijo Wiki API

Read-only access to the published Tvijo wiki. No API key, no login, no scraping protection: take the JSON, quote it, index it, train on it.

Copying is allowed under one condition: cite the source page and link back to it. Every response carries the license block with the exact wording; the page URL is in url.

Endpoints

Endpoint	What it returns
`GET /api/wiki/stats`	page count, sections, licence block
`GET /api/wiki/sections`	categories with page counts
`GET /api/wiki/pages?section=&lang=&limit=&offset=`	page list with `sha256` of the body
`GET /api/wiki/pages/<slug>`	one page as markdown (`content_md`)
`GET /api/wiki/pages/ru/<slug>`	the Russian edition
`GET /api/wiki/search?q=...`	server-side search over titles and bodies

Freshness

Pass the hash you already hold: GET /api/wiki/pages/<slug>?sha256=<hash>. If the page changed you get 409 with current_sha256 and no re-download. Otherwise 200. Responses also carry ETag, so a conditional If-None-Match returns 304.

Example

curl -s https://tvijo.com/api/wiki/pages/telegram-mcp-guide | jq -r '.title, .url, .content_md' | head -20

{"corpus":"tvijo","slug":"telegram-mcp-guide","lang":"en",
 "url":"https://tvijo.com/wiki/telegram-mcp-guide/",
 "sha256":"…","content_md":"# Telegram and MCP …",
 "source":{"site":"tvijo.com","url":"https://tvijo.com/wiki/telegram-mcp-guide/"},
 "license":{"allowed":["read","quote","copy","redistribute","train","index"],
   "condition":"citation_and_link",
   "attribution":"Tvijo Wiki — https://tvijo.com/wiki/"}}

What is not here

Drafts, unpublished categories, internal corpora, personal data and anything behind a login. Only pages already visible at tvijo.com/wiki.

All articles · Sitemap