Readhn

io.github.xodn348Published by GitHub account xodn348v0.2.0Updated Oct 24, 2026

AI-native HackerNews MCP Server. Find HN content that matters with explainable quality signals.

VerifiedSTDIODesktop onlyWeb Search & ScrapingData & Analytics

Overview

AI-generated overview

A local MCP server that lets an AI assistant search Hacker News, rank stories by explainable quality signals, and identify domain experts in comment threads.

What you can do
Discover and search stories: filter by keyword, score and time window, or run Algolia search with ranked results and quality signals per item. Analyze threads: pull a story with its top comments, or a full comment tree with per-comment signals. Profile people: find domain experts via EigenTrust propagation from seed experts and get a user profile with activity and trust score. Explain results: every response includes a breakdown of five signals (practitioner depth, thread depth, expert involvement, velocity, references).
When to use it
Useful when you want an assistant to answer questions about Hacker News discussion — top stories on a topic, what practitioners said in a thread, or who writes credibly about a field — with reasons for each ranking rather than a plain list.
Requirements
Runs locally as a Python package installed with pip, over stdio. No authentication or API keys are declared. Optional environment variables tune defaults: HN_KEYWORDS, HN_MIN_SCORE, HN_EXPERTS and HN_TIME_HOURS. A setup command can write configuration into supported AI agent clients. Network access to Hacker News is implied.
Before you install
The setup command modifies configuration files of detected AI agents; preview changes with the dry-run flag first. Queries and fetched Hacker News content are sent to and retrieved from third-party services (Hacker News and Algolia search). Ranking signals are heuristics, not verified quality judgments.

Installation

In SourceWeft

  1. Open Readhn in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.

Other MCP clients

Follow the launch instructions in the repository.

README

readhn

[PyPI] [Tests] [Coverage] [MCP]

AI-native HackerNews MCP Server. Find HN content that matters with explainable quality signals.

What It Does

Discover — Filter stories by keywords, scores, time. Get ranked results with quality signals.

Trust — Find domain experts. See who's talking and why they matter. EigenTrust propagation from seed experts.

Understand — Every result explains WHY. 5 signals: practitioner depth (30%), thread depth (20%), expert involvement (20%), velocity (15%), references (15%).

Quick Start

bash
# Installpip install readhn
# Auto-configure supported AI agentsreadhn setup

readhn setup detects Claude Code, Codex, Cursor, Claude Desktop, Cline, Windsurf, and OpenCode config paths and adds the readhn MCP server.

Useful setup flags:

bash
readhn setup --list              # Show detected agentsreadhn setup --dry-run           # Preview config changes onlyreadhn setup --agents "Cursor"  # Configure only specific agents

After setup, your AI agent auto-discovers readhn and uses it when you ask HN questions.

Usage

Ask your AI agent:

  • "Show me top HN stories about Rust this week"
  • "Find experts who write about databases on HN"
  • "What did practitioners say about Kubernetes networking?"

The agent calls readhn tools, gets results with quality signals, and explains why each result matters.

Configuration (Optional)

bash
export HN_KEYWORDS="ai,llm,rust,distributed-systems,databases"  # Default filter keywordsexport HN_MIN_SCORE="50"                                         # Minimum story scoreexport HN_EXPERTS="tptacek,simonw,antirez,ept,jepsen"           # Seed experts for trustexport HN_TIME_HOURS="24"                                        # Time window

How It Works

When you ask HN questions, your AI agent uses these tools:

  • discover_stories() — Top stories filtered by keywords/score/time, ranked by quality signals
  • search() — Algolia search with explainable ranking
  • find_experts() — Find domain experts using EigenTrust on comment graph
  • expert_brief() — User profile + activity + trust score
  • story_brief() — Story + top comments + signals in one call
  • thread_analysis() — Full comment tree with quality signals per comment

Every response includes signals breakdown: why each result was chosen.

License

MIT

Source: README.md at commit eb09fb5

Tools

0
SourceWeft has not started this package in its sandbox yet, so its tools are not known.

Version history

1
  1. v0.2.0LatestSep 16, 2026

Score

50% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 days25 pts

    Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.

  • Auth declared10 pts

    Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.

  • Tools listed15 pts

    Answer tools/list without a credential, or ship a package that starts in a sandbox without one.

  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts