When should you use it?
- You subscribe to many newsletters on one topic
- You would rather listen than read
- A computer can be awake early each day
Not for: Content you are not allowed to redistribute; Public feeds; keep the podcast private
What do you need, and what do you get?
You need
- Mail label collecting the newsletters
- Local text-to-speech tool
- Private podcast feed host
You get
- Daily written synthesis
- Audio episode in your podcast app
How does it work, step by step?
- Label the sourcesFilter the newsletters into one mail label.
- Start with textAsk for a written synthesis of the label to see if it is useful before adding audio.
- Add the voiceConvert the synthesis to speech with a local tool.
- Publish privatelyPost the episode to a feed only you subscribe to.
- Schedule itRun before you wake, with the computer set to wake itself.
The prompt
Our own prompt, inspired by Bronson Elliott and written to work in Claude Cowork and n8n. Values in lime are placeholders.
Every weekday at 5 AM, read all emails from the last 24 hours under my [label] label. Group stories that appear in several newsletters, drop repeats and keep only what is new. Write a five-minute conversational briefing naming which newsletter each story came from. Save it as text, convert it to audio with the local speech tool and publish it to my private podcast feed. If no new mail arrived, skip the episode.
Inspired by Bronson Elliott: I Built a Daily AI Podcast That Runs While I Sleep. The prompt above is written by playbooks.bot in our own words; nothing from the original is quoted.
What can go wrong?
- Keep the feed private; republishing others' newsletters publicly is a rights problem.
- A sleeping computer misses the run; set a wake schedule.
Our take
We like the suggested path more than the end state: start with a five-minute written summary of a label, and only build the audio chain if the summary earns it. Bronson's real problem was thirty-plus newsletters repeating the same stories, and deduplication is the part that earns its keep. A written synthesis of one label will tell you within a week whether the idea works for you. If it does, adding local text-to-speech and a private feed is a modest extra step. Keep that feed private; republishing other people's newsletters publicly is a rights problem, not a technical one. A computer that sleeps through the scheduled run produces nothing, so set it to wake before the job starts. Keep links to the originals in the written version so you can follow up on anything surprising. And use the briefing as a reason to unsubscribe: if a newsletter never adds anything the others missed, the inbox is better without it.
Where the time goes back: An hour of morning skimming becomes a single episode you hear while doing something else.
FAQ
Does it need a server?
Where should I start?
Source and credit
Inspired by Bronson Elliott. We describe the idea in our own words, link to the original, and wrote the prompt ourselves; nothing is quoted.
- Inspired by
- I Built a Daily AI Podcast That Runs While I Sleep
- Creator
- Bronson Elliott
- Published on
- bronsonelliott.substack.com (Substack)
- Original date
- Apr 14, 2026
- License
- Not openly licensed (the poster keeps all rights), so we quote nothing and link to the original
- Originally for
- Claude Cowork
- AI dry run
- Oct 23, 2026: passed. Dry-run tested by our AI test agent on sample input. PASS: Gates held on desk check; no live sends or purchases.
AI dry run on sample data; not run with live connected accounts. The playbooks.bot test agent filled the prompt with synthetic inputs, produced the output and checked it against the page's stated outputs and guardrails. - Verified
- Oct 24, 2026: source link resolves to the original, the credited creator published it, our excerpt or prompt matches it, and the license allows our use.
Are you the creator? Claim or correct this page.