Integration
OpenClaw + Superlore
Your AI agent already knows what you did today. Now hear it.
Give your OpenClaw agent a private podcast feed. It writes the briefing, narrates it locally, and publishes finished audio - you just press play in the podcast app you already use.
Why OpenClaw + Superlore?
Other audio tools need you to feed them content. Your OpenClaw agent already has the context - memory files, daily logs, project status, metrics. The Superlore skill simply turns what your agent already knows into a private show. Superlore never sees your workspace: the agent writes and narrates on your machine and uploads only the finished audio.
Zero manual input
OpenClaw already has the context - memory files, daily logs, project status. The skill turns what it knows into audio without you feeding it anything.
The agent does the audio
Narration runs locally on your machine with free Microsoft Edge TTS voices. Superlore hosts the feed and serves podcast apps the RSS they expect.
Private by default
Listen and publish tokens are separate, rotatable, and stored only as hashes. The QR subscribe page never exposes the publish token.
You can tell it is working
Reader counts, RSS reads, and publish activity show up on the feed dashboard - including failed uploads with error codes.
How it works
1. Create a private feed
Name it like a show ("Daily Agent Briefings"). Superlore generates cover art, a private RSS feed, and two separate tokens: one for listening, one for publishing.
2. Subscribe on your phone
Scan the QR code from the setup page and add the show to Apple Podcasts, Overcast, Pocket Casts, or any RSS-capable app. One time, then it is just a podcast.
3. Paste the skill prompt into OpenClaw
Copy the skill prompt from the setup page and send it to your agent in any channel. It stores a durable publishing workflow with the upload URL and publish token.
4. Ask for episodes - or schedule them
"Give me a podcast briefing on today's work." OpenClaw writes the script from what it already knows, narrates it locally, uploads the MP3, and the episode appears in your queue.
Set it up in one message
Create a feed, then copy the skill prompt from the setup page and send it to your agent. It looks like this (yours includes your real publishing instructions URL):
Install this as a durable agent skill for my private Superlore podcast feed.
Skill name: Superlore Agent Feed Publisher
Purpose: when I ask for a podcast episode or audio briefing, create
finished audio and publish it to my private Superlore feed.
First, read and store these publishing instructions:
https://superlore-api.onrender.com/agent/publish/feed_...?token=slp_...
Durable behavior:
1. Save this as a reusable skill, not a one-time chat note.
2. Use Microsoft Edge TTS with this voice: en-US-AndrewNeural.
3. Write a clear, listenable podcast script.
4. Generate an MP3 file.
5. Publish the MP3 to my Superlore feed using the instructions above.
...Make it a daily show
OpenClaw agents can run scheduled jobs. Once the skill is installed, ask for a standing schedule and your briefing becomes a real daily podcast:
Every weekday at 5pm, produce a podcast episode summarizing what you
worked on today, decisions made, and what's planned for tomorrow.
Publish it to my Superlore feed using the Superlore Agent Feed
Publisher skill.What the agent runs under the hood
# 1. Narrate the script it wrote (free, local)
edge-tts -f script.txt -v en-US-AndrewNeural --write-media episode.mp3
# 2. Publish the finished audio
curl -X POST "https://superlore-api.onrender.com/agent-feeds/<feedId>/items" \
-H "Authorization: Bearer $PUBLISH_TOKEN" \
-F "[email protected];type=audio/mpeg" \
-F "title=Daily Agent Briefing"Give your agent a voice today
Free accounts publish real feeds - 3 private feeds, 30 episodes each. Setup takes about two minutes end to end.
Questions? support at superlore.ai