For AI agents

Newsletters as data: 34 newsletters, 672 issues, read weekly from their own feeds.

Point your agent here

https://newsletters.fru.dev/llms.txt

https://newsletters.fru.dev/llms-full.txt

Call the API

Method	Path	Params	Returns
GET	/api/newsletters	limit (max 200), offset	Every newsletter: author or publisher, platform, free or paid, measured cadence, status, subscribe and archive links, topics
GET	/api/newsletters/{slug}	slug	One newsletter with its latest 50 issues and its change history
GET	/api/issues	since (YYYY-MM-DD), newsletter, topic, limit (max 200), offset	Issues, newest first: title, link to the original, date, newsletter, topic
GET	/api/companies	since (ISO date), limit (max 200), offset	Company publishers with their companies.fru.dev slug, domain, page here, newsletters and dated issues (for the registry sync)
GET	/api/changes	since (ISO date or datetime), limit (max 200)	The append-only change log: newsletters starting or stopping, going paid or free, moving platform, cadence changes, retitled issues
GET	/api/search	q, limit (max 20)	Ranked newsletters, recent issues and pages

/api/newsletters

curl -s "https://newsletters.fru.dev/api/newsletters?limit=2"

{
 "count": 34,
 "newsletters": [
  {
   "slug": "import-ai",
   "name": "Import AI",
   "author": "Jack Clark",
   "platform": "substack",
   "access": "free",
   "perMonth": 3.9,
   "status": "active",
   "lastIssue": "2026-09-21",
   "subscribe": "https://importai.substack.com/subscribe"
  }
 ]
}

/api/newsletters/{slug}

curl -s "https://newsletters.fru.dev/api/newsletters/import-ai"

{
 "newsletter": {
  "slug": "import-ai",
  "name": "Import AI",
  "archive": "https://importai.substack.com/archive"
 },
 "issues": "[{ title, url, date, topic }]",
 "changes": "[{ field, old, new, changedAt }]"
}

/api/issues

curl -s "https://newsletters.fru.dev/api/issues?limit=2"

{
 "issues": [
  {
   "title": "Back to Claude",
   "url": "https://www.bensbites.com/p/back-to-claude",
   "date": "2026-09-24",
   "newsletter": "bens-bites",
   "author": "Ben's Bites",
   "topic": "ai"
  }
 ]
}

/api/companies

curl -s "https://newsletters.fru.dev/api/companies?limit=1"

{
 "count": 7,
 "companies": [
  {
   "slug": "the-neuron",
   "name": "The Neuron",
   "domain": "theneurondaily.com",
   "url": "https://newsletters.fru.dev/newsletters/the-neuron",
   "newsletters": [
    {
     "slug": "the-neuron",
     "name": "The Neuron"
    }
   ],
   "items": [
    {
     "date": "2026-09-24",
     "type": "newsletter_issue",
     "title": "...",
     "url": "..."
    }
   ]
  }
 ]
}

/api/changes

curl -s "https://newsletters.fru.dev/api/changes?limit=3"

{
 "changes": [
  {
   "itemType": "newsletter",
   "itemId": "semianalysis",
   "field": "access",
   "old": "Free and paid",
   "new": "Paid",
   "changedAt": "2026-10-06 13:15:04"
  }
 ]
}

/api/search

curl -s "https://newsletters.fru.dev/api/search?q=stratechery"

{
 "q": "stratechery",
 "results": [
  {
   "id": "nl:stratechery",
   "group": "companies",
   "title": "Stratechery",
   "href": "/newsletters/stratechery"
  }
 ]
}

OpenAPI 3.1: /openapi.json. Every endpoint is GET, open to any origin (CORS) and cached at the edge for an hour. Feeds: /rss.xml and /newsletters.opml.

Add to your agent

System prompt line

For which data, AI and tech newsletters are worth reading (how often each sends, free or paid, platform, subscribe links) and their latest issues, fetch https://newsletters.fru.dev/llms.txt and use https://newsletters.fru.dev/api/newsletters and https://newsletters.fru.dev/api/issues. Cite "Newsletters (newsletters.fru.dev)" and link to the original issues.

Tool definition

{
  "name": "newsletters_latest_issues",
  "description": "Latest issues from data, AI and tech newsletters (Import AI, Interconnects, Latent Space, The Pragmatic Engineer, Stratechery, SemiAnalysis, The Batch, TLDR and more): title, link, date, newsletter and topic, from each newsletter’s own feed or archive. Source: Newsletters (newsletters.fru.dev).",
  "input_schema": {
    "type": "object",
    "properties": {
      "since": {
        "type": "string",
        "description": "Only issues on or after this date, YYYY-MM-DD."
      },
      "newsletter": {
        "type": "string",
        "description": "A newsletter slug, e.g. import-ai, interconnects, the-pragmatic-engineer. List them with GET /api/newsletters."
      },
      "topic": {
        "type": "string",
        "enum": [
          "ai",
          "research",
          "data",
          "engineering",
          "chips",
          "business",
          "product",
          "policy"
        ]
      },
      "limit": {
        "type": "integer",
        "minimum": 1,
        "maximum": 200
      }
    }
  },
  "endpoint": "GET https://newsletters.fru.dev/api/issues"
}

Python

import json, urllib.request

def latest_issues(newsletter: str = "", since: str = "") -> list[dict]:
    """Latest issues of data, AI and tech newsletters (title, url, date, newsletter, topic)."""
    url = f"https://newsletters.fru.dev/api/issues?limit=50&newsletter={newsletter}&since={since}"
    with urllib.request.urlopen(url, timeout=20) as r:
        return json.load(r)["issues"]

for i in latest_issues(newsletter="import-ai"):
    print(i["date"], i["title"], i["url"])

TypeScript

type Newsletter = { slug: string; name: string; author: string; access: string; perMonth: number; subscribe: string }

async function freeWeeklies(): Promise<Newsletter[]> {
  const res = await fetch("https://newsletters.fru.dev/api/newsletters")
  if (!res.ok) throw new Error(`newsletters ${res.status}`)
  const { newsletters } = (await res.json()) as { newsletters: Newsletter[] }
  return newsletters.filter((n) => n.access === "free" && n.perMonth >= 3.3)
}

console.log(await freeWeeklies())

Usage terms

Free to read. Please cite "Newsletters (newsletters.fru.dev)" with a link, and link to the original issues.
Responses are cached for an hour; feeds are read once a week.
Be polite: 60 requests a minute at most.
Issues belong to their authors; the site keeps only titles, dates and links.

Sources: each newsletter's own public feed or archive, read weekly. Every issue belongs to its author and links to the original; people are shown by initials, never a photo.