eksisozluk-mcp
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@eksisozluk-mcp What's trending on ekşi sözlük right now?"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
eksisozluk-mcp 🇹🇷
mcp-name: io.github.mohammad-abbasi-sh/eksisozluk-mcp
Unofficial read-only MCP server for eksisozluk.com — Ekşi Sözlük, Türkiye's iconic collaborative dictionary and social platform ("kutsal bilgi kaynağı"), where users write short entries under topic titles. Trending topics (gündem), topic search, entries with şükela sorting, channels, and yesterday's most-liked entries (debe).
Built by Mohammad Abbasi as part of the "Turkey MCP factory" (Sepehr-style pipeline). Not affiliated with eksisozluk.com or Ekşi Sözlük.
Content note: entries are unfiltered user-generated text — they may contain profanity, slang and strong opinions. The server returns what the site publishes, verbatim; nothing is censored or rewritten.
For AI agents
Add to your MCP client config (stdio):
{
"mcpServers": {
"eksisozluk": {
"command": "uvx",
"args": ["eksisozluk-mcp==0.1.0"]
}
}
}Or run locally: python -m eksisozluk_mcp.server (needs mcp<2, curl-cffi).
Related MCP server: Product Hunt MCP
Tools
Tool | Description |
| Gündem — trending topics right now; |
| Search topic titles (partial matches welcome) — |
| Entries of one topic — |
| Topics of one channel — |
| Yesterday's most-liked entries (debe), grouped by topic; |
| Check whether eksisozluk.com is reachable (ok / blocked / error) |
Example output (eksisozluk_get_topic / ankara, 2026-11-12):
{"topic": "ankara", "topic_id": "32788", "sort": "recent",
"page": 1, "page_count": 3621,
"entries": [{"entry_id": "3820", "author": "zomzom",
"date": "28.03.1999 ~ 20.11.2005 22:40",
"text": "yapay kent. baskent olmasaydi boktan bi yer olacakti muhakkak.",
"url": "https://eksisozluk.com/entry/3820"}]}How it works
eksisozluk.com serves its pages fully server-rendered — no login, no API key, no session cookies. Topic lists are <ul class="topic-list partial"> rows (<a href="/<slug>--<id>">title <small>count</small></a>, 50 per page, ?p=2 pagination). Topic pages list 10 entries per page in <ul id="entry-item-list"> (<li data-id> with a <div class="content">, author link /biri/<nick>, date link /entry/<id>); the pager carries data-pagecount (ankara: 3,621 pages; starbucks: 888). Sort modes are ?a=popular (şükela), ?a=dailynice (son 24 saat), ?a=nice (şükela tümü). The search box's exact-title lookup (/?q=<title>) 302-redirects to the topic only for an exact title and 404s otherwise — partial matches go through /basliklar/ara?SearchText=. Debe (/debe) groups yesterday's most-liked entries by topic, each headed by an <h1 id="title">. Transport uses Chrome TLS impersonation (curl-cffi, chrome120 with rotation); requests are spaced ~0.8s apart; everything is read-only.
Install
uvx eksisozluk-mcp==0.1.0 # run directly (needs uv)
# or
pip install eksisozluk-mcp==0.1.0
eksisozluk-mcp # starts the MCP server on stdioTests
python -m venv .venv && .venv/bin/pip install -e .
.venv/bin/python -m unittest discover -s tests -vOffline tests run on recorded fixtures (real pages fetched 2026-11-12); live tests hit eksisozluk.com. Tests green at build time (see repo for counts) plus an MCP stdio smoke test.
Roadmap
Author pages (
/biri/<nick>— an author's recent entries)Entry comments (yorum) under an entry
"tarihte bugün" (on this day) topics
License
MIT — see LICENSE.
Built by Mohammad Abbasi — Almanibama · mabbasi.ir
Available Tools
6 toolseksisozluk_channel_topicsB
Topics of one eksisozluk.com channel.
channel: e.g. "spor", "siyaset", "iliskiler", "yasam", "seyahat", "muzik", "tv", "haber", "bilim", "edebiyat", "kripto". page: list page (50 topics per page). limit: 1..50 topics returned.
Inputs: channel (required), page, limit. Returns: result
eksisozluk_debeA
Yesterday's most-liked entries (debe), grouped by topic.
limit: 1..12 topic groups returned (the site lists ~10-12). Each group: topic_slug, topic_id, topic_title, topic_url, entries (usually 1 top entry per topic).
Inputs: limit. Returns: result
eksisozluk_get_topicA
Entries of one topic.
topic: "slug--id" (e.g. "ankara--32788"), or an EXACT topic title (e.g. "ankara" — resolved via the site's /?q= redirect; a bare numeric id alone is not enough, and partial titles must go through eksisozluk_search_topics first). sort: "recent" (default), "popular" (şükela), "dailynice" (son 24 saat), "nice" (şükela tümü). page: entries page (10 per page); page_count is returned. limit: 1..10 entries returned. Each entry: entry_id, text, author, author_url, date, url.
Inputs: topic (required), page, sort, limit. Returns: result
eksisozluk_healthA
Check whether eksisozluk.com is reachable right now.
Returns status ok (HTTP 200), blocked (HTTP 403/429 — bot wall or rate limit) or error, plus the HTTP status and the probed URL. One polite request; it does not fetch any listings.
Inputs: no named parameters. Returns: result
eksisozluk_search_topicsA
Search topic titles on eksisozluk.com (partial matches welcome).
query: Turkish search text (e.g. "ankar"). page: results page (50 topics per page). limit: 1..50 topics returned. Each topic: title, slug, topic_id, entry_count, url.
Inputs: query (required), page, limit. Returns: result
eksisozluk_trendingA
Gündem — the trending topics on eksisozluk.com right now.
page: list page (50 topics per page). limit: 1..50 topics returned. Each topic: title, slug, topic_id, entry_count, url.
Inputs: page, limit. Returns: result
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 6 tools
Each tool targets a clearly distinct resource or action: channel topic listing, trending topics, title search, full topic entries, daily best (debe), and a health probe. The only mild adjacency is between channel_topics and trending, but their descriptions make the boundary explicit.
All names share the eksisozluk_ prefix and snake_case formatting, so the namespace is predictable. However, several names are bare noun phrases (debe, health, trending, channel_topics) while others use verb_noun (search_topics, get_topic), so the action pattern is not fully uniform.
Six tools is well-scoped for a read-only site browser, comfortably within the 3–15 range. Each tool covers a distinct browsing need and none feels like filler.
The surface covers discovery (trending, channel, search, debe), content retrieval (get_topic), and connectivity (health), which supports the core workflows. Minor gaps remain: no author/user profile browsing and no direct single-entry or random-topic fetch, though agents can work around these via get_topic.
Maintenance
Related MCP Connectors
Read-only MCP for the Eco game wiki: search, Markdown pages, and wiki_* lookups. No keys, no writes.
Read-only access to InfluSense influencer discovery, ratings, watchlists, and reports via MCP.
Search MCP servers, MCP clients and AI agents, and retrieve listing details. Free, read-only access.
Read-only MCP tools for AI agent discovery, structured resources, and NIULAI information.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceEnables MCP clients to scan, search, and read Reddit posts and comments using Reddit's public JSON endpoints, with no API credentials required.34 npmMIT
- AlicenseNot gradedqualityBmaintenanceEnables read-only access to Product Hunt data, including launches, post details, comments, topic search, and user profiles, through MCP tools over Streamable HTTP.MIT
- AlicenseNot gradedqualityBmaintenanceEnables read-only exploration of community content such as posts, topics, games, comments, user profiles, notifications, discounts, news, and trending items through MCP tools or a CLI.MIT
- AlicenseBqualityBmaintenanceEnables read-only access to BuzzerBeater's public API through MCP tools for retrieving team, player, match, market, national team, and other game data.29MIT