Skip to templates
AI Craft Pad
Menu
CONNECTED SERVICE RECIPE · Content

RSS digest → Telegram

Read an approved feed and send up to five dated HTTPS items from the last 24 hours.

Preview by defaultOwn accounts required for live modeMIT · AI Craft Pad
1Read approved feed
2Filter dates and links
3Build one digest
4Send to chosen chat

What happens in live mode

Live mode uses n8n RSS Read with TLS verification enabled, checks publication dates against the current time, removes repeated links within the run and sends at most five recent items to your configured Telegram chat. Future, undated, stale and unsupported links are omitted. Preview uses a fixed synthetic time and feed.

The downloaded workflow is unpublished and its Configuration mode is preview. In preview, the service branch is skipped. The website demo below also stays local. Changing a JSON input cannot enable live mode; that setting belongs to the workflow owner.

LOCAL PREVIEW · NO SERVICE REQUESTS

Inspect the data before connecting.

Edit the synthetic input to see the proposed fields, rows or report text. This is a transformation preview, not a call to HubSpot, Google, a feed or Telegram. No credentials belong in this form.

Ready. Inspect the preview before any live configuration.

Connect and test in your own n8n

Import the file without publishing it. The workflow includes its source, validation, duplicate/empty handling where applicable, and destination nodes. Credentials are intentionally unbound; select your own credentials in n8n rather than editing secrets into Code nodes.

  1. Replace the example feed URL in Configuration with a public HTTPS feed you are authorized to read. Review that domain and its content; do not use internal URLs or authenticated feeds in this starter.
  2. Create an n8n Telegram API credential and select it in Send reviewed report. Start the bot in your intended test chat and set that exact chat_id.
  3. Run preview to inspect title, link and date handling. In the local preview as_of is explicit; live mode uses the actual run time.
  4. Set mode to live and run manually against the approved feed and test chat. Check that only the latest five eligible items appear and titles remain literal text.
  5. An empty or stale feed should send nothing. If adding a schedule, first add a durable sent-item ledger and verify repeat-run behavior; this starter only deduplicates within one run.
Download connected n8n recipe

SHA-256: d6ac8a040de9a354b55766e5092949adf9d321201cf9b951139e55eb7ffd4da6

Failure, repeat runs and recovery

No cross-run cursor or sent-item database is included, so repeating a run can resend items still inside the 24-hour window. The feed is read by your n8n instance; use a small trusted feed. The local selection considers at most 100 parsed entries, preserves a conservative HTTPS link format and caps the report at 3,900 characters.

HTTP request nodes have a 10-second timeout and stop on errors. Writes are not automatically retried. Authentication failures, rate limits and ambiguous timeouts require inspection in your own n8n execution view and destination account before rerunning. Do not turn a provider error into an empty successful lookup.

Data handling

The chosen public feed is fetched by your n8n instance and selected titles/links go to Telegram only in live mode. Article bodies are not summarized or sent. Source text remains untrusted data and never becomes code or setup instructions.

The export disables saved success/error execution data and manual-execution saving. Your instance logs and provider retention still apply; review those settings for your account. No live account is connected to AI Craft Pad by this page.

Costs and requirements

Needs your n8n instance and a Telegram bot/chat. Hosting, feed-provider terms and Telegram limits apply; no paid AI API or paid broadcast is used.

Continue with another task