# Draft your weekly newsletter from the week’s best sources. News, Hacker News, X and Reddit, read for you

A newsletter draft picks the week’s best stories on your topic and gives a short take on each. With Agentik, your AI collects news with Serper, stories from Hacker News, posts on X and threads from Reddit through ScrapeCreators, then reads the best stories in full with Firecrawl.

- For: Newsletter writers, community and content teams
- Providers: 5 (Serper, Hacker News, X, ScrapeCreators and Firecrawl), 5 tools
- Example (3 news searches, 30 Hacker News stories, 50 X posts, one Reddit search, 8 stories read): at most $0.81, computed from the live catalog
- Updated: 2026-10-10

## Each provider does one part

No single provider covers this job. Through Agentik your AI calls each one for its part, from one connection and one prepaid balance.

- **Serper** (1 tool, at most $0.03): The week’s Google News results on your topic.
- **Hacker News** (1 tool, free): Hacker News stories and their discussion.
- **X** (1 tool, at most $0.75): What people post about the topic this week.
- **ScrapeCreators** (1 tool, at most $0.0019): Reddit threads from the week.
- **Firecrawl** (1 tool, at most $0.02): The best stories read in full, so each take is accurate.

## Give this to your agent

```text
Use my connected Agentik MCP server for this job: weekly newsletter draft.

Draft this week’s newsletter. First ask me for my topic, my audience, my usual sections and two past issues for my voice.

Collect the week’s news with serper/news, Hacker News stories with hackernews/search, X posts with x/search-recent and Reddit threads with scrapecreators/get-v1-reddit-search. Read the eight best stories in full with firecrawl/scrape.

Finish with a draft in my sections and voice (headline, intro, five items with a two-line take and a link each, one thread worth reading, one thing to watch).

Before running anything, inspect each tool with inspect_tool, ask me for any input you don’t have instead of inventing it, and tell me the total price. Wait for my go. At the end, tell me what Agentik charged.
```

## You provide

- Your topic
- Your audience and sections
- Two past issues

## What you get: a newsletter draft with links

- The week’s best stories from four sources
- A two-line take on each, with a link
- A thread worth reading
- A full draft in your voice

## How to draft a weekly newsletter in Claude or ChatGPT

Prices are the most each call can reserve for this example: 3 news searches, 30 Hacker News stories, 50 X posts, one Reddit search, 8 stories read.

### 1. Collect the week

Serper, Hacker News, the X API and ScrapeCreators collect the week’s news, stories, posts and Reddit threads on your topic.

- [Google News search](/tools/serper/news.md) `serper/news` (Serper, 3 searches): at most $0.03
- [Hacker News search](/tools/hackernews/search.md) `hackernews/search` (Hacker News, 30 stories): free
- [X recent search](/tools/x/search-recent.md) `x/search-recent` (X, 50 posts): at most $0.75
- [Reddit search](/tools/scrapecreators/get-v1-reddit-search.md) `scrapecreators/get-v1-reddit-search` (ScrapeCreators, 1 search): at most $0.0019

### 2. Read the best stories

Firecrawl reads the eight best stories in full, so each take is accurate.

- [Scrape a web page to markdown](/tools/firecrawl/scrape.md) `firecrawl/scrape` (Firecrawl, 8 stories): at most $0.02

### 3. Write the draft

Your AI writes the draft in your sections and voice, with a link for each item.

Example total: at most $0.81. Each call is billed for measured usage; unused reservations are released.

## Ask it your way

- Draft this week’s newsletter about AI agents.
- What were the most discussed stories on my topic this week?
- Write a two-line take on each of the week’s top stories.
- Find one Reddit or Hacker News thread worth sharing this week.

## Why through Agentik

- News, Hacker News, X, Reddit and page reading come from five providers through one connection.
- Hacker News is free; X posts are most of the price.

## Good to know

- Every item links to its source; your AI never writes a take on a story it did not read.
- X posts are billed per post returned; Hacker News is free.

## When to use something else

To send, segment and track a newsletter, use your email platform. This playbook drafts the issue from the week’s sources.

## Weekly newsletter draft: questions

### Will it write about stories it didn’t read?

No. Every item links to a source it read in full.

### Can it match my voice?

Give your agent two past issues and it follows their tone and sections.

### What does the example cost?

At most $0.81, reserved before each call and settled at the measured usage.

## Keep going

- [Audio news briefing](/audio-news-briefing.md): Today’s news on your topic, read in full, turned into a two-host audio briefing with a short jingle.
- [LinkedIn post with sources](/linkedin-post-engine.md): The LinkedIn posts that did best on your topic this month, fresh data to cite from the last weeks, then a post in your voice with a visual.
- [Social listening](/social-listening.md): Posts on X, Hacker News threads, YouTube comments and news coverage, read together and sorted into themes.
- [News tools for your agent](/tools/news.md)
- [Social tools for your agent](/tools/social.md)
- [Scraping tools for your agent](/tools/scraping.md)

[All playbooks](/playbooks.md) · [Connection guides](/docs/clients.md) · [Billing](/docs/billing.md)

---

[HTML page](/newsletter-draft) · [Agent index](/llms.txt)
