Agent skill

hn

View SKILL.md on GitHub Repository
Stars 560
Forks 10

Install this agent skill to your Project

npx add-skill https://github.com/nexi-lab/nexus/tree/develop/src/nexus/backends/connectors/hn

SKILL.md

Hacker News Connector

Mount Path

/mnt/hn/

Overview

The Hacker News connector provides read-only filesystem access to Hacker News stories and comments. Stories are organized by feed type (top, new, best, ask, show, jobs) and represented as JSON files with nested comments.

No authentication required - uses the public HN Firebase API.

Directory Structure

/mnt/hn/
  top/                    # Top-ranked stories
    1.json ... 10.json    # Stories ranked 1-10
  new/                    # Newest submissions
    1.json ... 10.json
  best/                   # Algorithmically best stories
    1.json ... 10.json
  ask/                    # Ask HN posts (community questions)
    1.json ... 10.json
  show/                   # Show HN posts (project showcases)
    1.json ... 10.json
  jobs/                   # Job postings
    1.json ... 10.json

Operations

List Feeds

List available feed categories:

bash
nexus ls /mnt/hn/

Returns:

top/
new/
best/
ask/
show/
jobs/

List Stories in Feed

List stories in a specific feed:

bash
nexus ls /mnt/hn/top/

Returns:

1.json
2.json
3.json
...
10.json

Read Story

Read a story with its comments:

bash
nexus cat /mnt/hn/top/1.json

Returns JSON with story metadata and nested comments:

json
{
  "id": 12345678,
  "type": "story",
  "by": "username",
  "time": 1704067200,
  "title": "Show HN: My new project",
  "url": "https://example.com/project",
  "score": 142,
  "descendants": 87,
  "kids": [12345679, 12345680],
  "_rank": 1,
  "_feed": "top",
  "comments": [
    {
      "id": 12345679,
      "type": "comment",
      "by": "commenter1",
      "time": 1704067500,
      "text": "Great project! I especially like...",
      "parent": 12345678,
      "replies": [
        {
          "id": 12345681,
          "type": "comment",
          "by": "author",
          "text": "Thanks! We spent a lot of time on...",
          "replies": []
        }
      ]
    }
  ]
}

Search Stories

Search across cached stories:

bash
nexus grep "machine learning" /mnt/hn/top/
nexus grep "Show HN" /mnt/hn/

Feed Types

Feed Description Update Frequency
top Top-ranked stories by score ~5 minutes
new Newest submissions ~1 minute
best Algorithmically selected best ~1 hour
ask "Ask HN" community questions ~5 minutes
show "Show HN" project showcases ~5 minutes
jobs Y Combinator job postings ~1 hour

Story Fields

Field Type Description
id int Unique story identifier
type string "story", "job", "poll"
by string Author username
time int Unix timestamp
title string Story title
url string Link URL (if link post)
text string Post text (if text post)
score int Upvote score
descendants int Total comment count
kids int[] Direct child comment IDs
_rank int Position in feed (1-10)
_feed string Feed name (top/new/etc)
comments array Nested comment objects

Comment Fields

Field Type Description
id int Unique comment identifier
type string Always "comment"
by string Author username
time int Unix timestamp
text string Comment text (HTML)
parent int Parent story/comment ID
dead bool True if flagged/dead
deleted bool True if deleted
replies array Nested reply comments

Examples

Get Today's Top Story

bash
nexus cat /mnt/hn/top/1.json | jq '.title, .url, .score'

Find AI-Related Stories

bash
nexus grep -i "artificial intelligence\|AI\|GPT\|LLM" /mnt/hn/top/

Get Ask HN Discussions

bash
nexus cat /mnt/hn/ask/1.json | jq '.title, .text'

Browse Job Postings

bash
for i in 1 2 3; do
  nexus cat /mnt/hn/jobs/$i.json | jq -r '.title'
done

Find Stories by Author

bash
nexus grep '"by": "dang"' /mnt/hn/

Caching

Stories are cached with TTL based on feed type:

Feed Cache TTL Reason
new 1 minute Changes very frequently
top 5 minutes Moderately dynamic
ask 5 minutes Moderately dynamic
show 5 minutes Moderately dynamic
best 1 hour Relatively stable
jobs 1 hour Changes slowly

Use sync to pre-fetch and cache all stories:

bash
nexus sync /mnt/hn/
nexus sync /mnt/hn/top/  # Sync specific feed

Limitations

  • Read-only: HN API does not support posting or voting
  • 10 stories per feed: Configurable up to 30 via connector settings
  • Comment depth limit: Max 5 levels deep, 100 comments total
  • No search API: Search is local (grep on cached content only)
  • No user profiles: Cannot fetch user details directly
  • External content not included: Only URLs, not article content

HN Terminology

Term Meaning
karma User reputation score from upvotes
Ask HN Questions posed to the community
Show HN Project/creation showcases
dead Flagged/hidden by moderators
dang HN moderator (Daniel Gackle)
pg Paul Graham (YC founder)
flagged Reported by users
[dupe] Duplicate submission

Rate Limits

The HN Firebase API has no documented rate limits, but the connector implements reasonable caching to avoid excessive requests.

Related Resources

Recommended Agent Skills

Expand your agent's capabilities with these related and highly-rated skills.

nexi-lab/nexus

herb-enterprise-context

Search enterprise context data (Slack, docs, meetings, PRs) from the HERB benchmark to answer questions about products, employees, customers, and organizational activities. Use when answering questions that require searching through enterprise communication and documentation.

560 10
Explore
nexi-lab/nexus

mcp-builder

Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).

560 10
Explore
nexi-lab/nexus

calendar

560 10
Explore
nexi-lab/nexus

gmail

560 10
Explore
microsoft/apm

auth

Activate when code touches token management, credential resolution, git auth flows, GITHUB_APM_PAT, ADO_APM_PAT, AuthResolver, HostInfo, AuthContext, or any remote host authentication — even if 'auth' isn't mentioned explicitly.

1,149 100
Explore
microsoft/apm

python-architecture

Activate when creating new modules, refactoring class hierarchies, introducing design patterns, or making changes spanning 3+ files in the APM CLI codebase.

1,149 100
Explore

Didn't find tool you were looking for?

Be as detailed as possible for better results