Hacker News MCP — by NeCL – Neural Engineering & Cognitive Logic

@adjacentai

リポジトリ

基本情報

カテゴリ

開発者ツール

ライセンス

MIT

トランスポート

stdio

作者

adjacentai

投稿者

necl_adjacentai

Hacker News MCP — by NeCL – Neural Engineering & Cognitive Logic について

Hacker News tools for AI agents (MCP server). Top stories, comments thread, full-text search via Algolia. No API key required. By NeCL.

#hacker-news #hn #research #news #ai-agents #content #claude #mcp#built-by-necl

Hacker News MCP — by NeCL – Neural Engineering & Cognitive Logic の設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

{
  "mcpServers": {
    "necl-hn": {
      "command": "uvx",
      "args": [
        "--from",
        "git+https://github.com/adjacentai/necl-hn-mcp.git",
        "necl-hn-mcp"
      ]
    }
  }
}

Hacker News MCP — by NeCL – Neural Engineering & Cognitive Logic のツール

Get the top Hacker News stories from the last N hours, ranked by score. Args: limit: Number of stories to return (1-30). Default 10. hours: Time window in hours back from now. Default 24. Returns: List of story dicts with id, title, url, score, author, comments_count, posted_unix, hn_url, type, text (for Ask/Show HN posts).

Get stories from a specific HN category. Args: category: One of "top", "new", "best", "ask", "show", "job". limit: Number of stories to return (1-30). Default 10. Returns: List of story dicts.

Get full metadata for a single HN story by id. Args: story_id: The HN item id. Returns: Story dict with id, title, url, score, author, comments_count, posted_unix, hn_url, type, text. Returns {"error": "..."} if not found.

Get the top-level comments thread for a story, in HN ranking order. Args: story_id: The HN item id of the parent story. limit: Number of comments to return (1-30). Default 10. Returns: List of comment dicts with id, author, text, posted_unix, replies_count.

Full-text search across HN stories and comments via Algolia. Args: query: Search query string. sort: "relevance" (default) or "date" (newest first). limit: Number of hits to return (1-50). Default 20. Returns: Dict with `hits` (list of clean dicts: id, kind, title, url, hn_url, author, points, comments_count, story_text, comment_text, created_at, tags) and `total_hits` (total matches Algolia found, may exceed limit).

概要

What is Hacker News MCP — by NeCL – Neural Engineering & Cognitive Logic?

A set of tools for AI agents to access Hacker News top stories, story details, comments threads, and full-text search via the public HN API. It requires no credentials, is zero‑config, and is built by NeCL – an AI engineering studio that uses it internally.

How to use Hacker News MCP — by NeCL – Neural Engineering & Cognitive Logic?

Install via uvx (recommended, no setup) or pip, then add the configuration to your MCP client’s mcp.json. Once configured, agents can call tools like hn_top_stories, hn_get_story, hn_get_comments, hn_search, and hn_category using natural language prompts.

Key features of Hacker News MCP — by NeCL – Neural Engineering & Cognitive Logic

Zero credentials and zero configuration required.
Five tools for top stories, story details, comments, search, and categories.
Battle‑tested code extracted from NeCL’s internal content engine.
Clean structured data output, designed for LLM chaining.
Compatible with any MCP client (Claude Desktop, Claude Code, etc.).
Uses the free public HN API with no rate limits worth worrying about.

Use cases of Hacker News MCP — by NeCL – Neural Engineering & Cognitive Logic

Generate daily content briefs from top HN stories.
Perform competitive research by searching HN for company mentions.
Monitor trends by polling new and best categories on a cron.
Build a newsletter pipeline that searches by keyword, clusters results, and creates editorial digests.
Automate customer‑discovery by searching Ask HN for problems your product solves.

よくある質問

`mcp` command not found or install errors?

Ensure Python 3.10+ is installed. If using `uvx`, install `uv` (`pip install uv`). For pip install, run `pip install necl-hn-mcp` and verify with `python -m necl_hn_mcp --help`.

Tool calls timeout?

HN’s Firebase API can be slow when fetching many items. For long‑window queries (e.g. `hn_top_stories(limit=30, hours=168)`), reduce the `limit` value. Default timeout is 10 seconds per request.

Algolia returns no hits for my search?

Algolia indexes HN content with a delay of about 5–15 minutes for fresh items. Use `sort="date"` to prioritise recent activity.

Why are comments threads empty for some stories?

`hn_get_comments` returns only top‑level comments to keep token costs predictable. The `replies_count` field on each comment indicates if deeper discussion exists, but replies are not recursively fetched.

Tools don’t show up in Claude after configuration?

Fully restart Claude Desktop or Claude Code (not just reload). Check the MCP logs under Settings → Developer to verify the server connects.

「開発者ツール」の他のコンテンツ

Modelglass

Modelglass turns "which model should I use?" into a lookup instead of a guess. Seven tools give your agent live pricing across image, language, video, and audio models. Sourced and versioned, not scraped once and forgott

NextBlock CMS

NextBlock

Open-source Next.js + Supabase CMS with a per-site MCP server: scaffold a site, then create pages, posts, navigation, media and SEO from Claude, Cursor or any MCP client. Streamable HTTP with a bearer token you mint in y

Vibgrate MCP

Vibgrate

Query your team's drift, vulnerability, and migration data from any AI assistant. Vibgrate MCP connects Cursor, Claude, ChatGPT, Windsurf, or VS Code to Vibgrate Cloud: 51 tools for DriftScores, CVEs and EOL runtimes, up

LastPing

Know when your AI agents, cron jobs and CI pipelines fail, stall or go quiet. Your agent sets up monitors from a plain description, tells you what is broken and why, reads run history and traces, and writes notes on inci

LocalCan

Gives AI agents public URLs (tunnels) for localhost, live HTTP traffic inspection, snapshot publishing, and access control. Part of LocalCan, the ngrok alternative for Mac, Windows and Linux. Free plan.