HackerNews Monitor + Auto-Tagger

Search and monitor HackerNews stories, Show HN, Ask HN, polls, jobs, comments by keyword and date range. Auto-tag every post with your custom keyword list. Built on the official Algolia public API — no auth, no rate-limit headaches, full historical archive.

What it does

• Multi-query search (run several keywords in one execution, results grouped by query)
• Filter by post type: story, comment, Show HN, Ask HN, poll, job, front_page
• Date range filter (any time window since HN inception)
• Sort by popularity (best matches) or date (newest)
• Min points / min comments thresholds
• Auto-tag every post with a keyword list — perfect for trend monitoring across multiple themes

Use cases

Tech market intelligence — Track every Show HN launch in your category to spot competitors and adjacent products.

AI trend detection — Monitor "RAG", "agent", "fine-tuning" to catch emerging topics before they go mainstream.

Founder lead-gen — Find Ask HN posts asking for tools in your space, reach out before they buy elsewhere.

LLM training data — Curate high-quality HN discussions as a corpus for fine-tuning domain-specific models.

Content marketing — Find high-engagement HN stories (>100 points, >50 comments) for blog post inspiration.

Input

{
  "queries": ["RAG", "LLM agents", "AI infrastructure"],
  "postType": "story",
  "sortBy": "popularity",
  "dateFrom": "2026-04-01",
  "dateTo": "2026-04-29",
  "minPoints": 5,
  "minComments": 0,
  "maxResultsPerQuery": 50,
  "tagKeywords": ["RAG", "Llama", "fine-tuning", "agent", "embedding", "vector-db", "evaluation"]
}

Field	Type	Default	Description
queries	array	required	Search keywords (one per query)
postType	enum	story	story / comment / show_hn / ask_hn / poll / job / front_page / all
sortBy	enum	popularity	popularity (best matches first) or date (newest first)
dateFrom	YYYY-MM-DD	(none)	Lower date bound (inclusive)
dateTo	YYYY-MM-DD	(none)	Upper date bound (inclusive)
minPoints	int	5	Skip posts below this karma
minComments	int	0	Skip posts below this comment count
maxResultsPerQuery	int (1-1000)	50	Cap per query
tagKeywords	array	[]	Keywords auto-tagged on title

Output

One dataset item per matching post:

{
  "query": "RAG",
  "objectID": "47955083",
  "type": "story",
  "title": "Production RAG: what I learned from processing 5M+ documents",
  "url": "https://example.com/blog/rag-lessons",
  "hnUrl": "https://news.ycombinator.com/item?id=47955083",
  "author": "user1",
  "points": 551,
  "numComments": 114,
  "createdAt": "2026-04-15T13:42:00.000Z",
  "createdAtEpoch": 1745000000,
  "storyId": null,
  "tags": ["RAG", "evaluation"]
}

Pricing

Currently FREE during the launch preview — no per-result charges, no monthly cap.

When paid pricing rolls out (notice will be posted at least 14 days in advance):

Event	Price
Actor start	$0.01 (one-time per run)
Result item	$0.001 (per post)

Cost examples (post-launch):

• 100 posts: ~$0.11
• 1.000 posts: ~$1.01
• 10.000 posts: ~$10.01

Limits

• Source: HackerNews Algolia public search API (no auth required)
• Coverage: full HN archive (2007 onwards)
• Rate limit: ~10 req/sec (politely paced internally with 0.3s sleep)
• Max results per query: 1.000 per run
• Comment text not included by default (use HN item URL for full thread)

Source attribution

Data comes from the HackerNews Algolia public API, maintained by Algolia in partnership with Y Combinator. Subject to YC's terms.

Author

Polara Data — niche scrapers for Italy, EU & global markets.