Podcast Production Pipeline
An agent chain that turns an episode topic into research, an outline, timestamped show notes, descriptions, and social media promo posts.
The pipeline covers two production stages: pre-recording (guest and topic research, episode outline) and post-recording (timestamped show notes, description, promo). Optionally, you can set up RSS monitoring of other podcasts. Pairs well with the Multi-Agent Content Factory to repurpose podcast content into blog posts, newsletters, or video clips.
Requirements
Web search / research skill — for guest research and topic deep-dives.
File system access — for reading transcripts and writing output files.
Slack, Discord, or Telegram integration — for delivering assets.
Optional: sessions_spawn for running research and writing agents in parallel.
Optional: RSS feed skill (for monitoring competitor podcasts).
>Как настроить
- Pre-recording, generate research and an outline: fill in the topic and guest in the prompt below, replacing the placeholders. Everything is saved to ~/podcast/episodes/[episode-number]/prep/.
I'm recording a podcast episode about [TOPIC]. My guest is [NAME].
Please:
1. Research the guest — their background, recent work, hot takes, and
anything controversial or interesting they've said publicly.
2. Research the topic — key trends, recent news, common misconceptions,
and what the audience likely already knows vs. what would surprise them.
3. Generate an episode outline:
- Cold open hook (1-2 sentences to grab attention)
- Intro script (30 seconds, casual tone)
- 5-7 interview questions, ordered from easy/rapport-building to deep/provocative
- 2-3 "back pocket" questions in case the conversation stalls
- Closing segment with call-to-action
Save everything to ~/podcast/episodes/[episode-number]/prep/
- Post-recording, generate show notes and promo: paste the transcript or point to its file in the prompt below. Everything is saved to ~/podcast/episodes/[episode-number]/publish/.
Here's the transcript for Episode [NUMBER]: [paste or point to file]
Please:
1. Write timestamped show notes — every major topic shift gets a timestamp
and one-line summary. Include links to anything mentioned (tools, books,
articles, people).
2. Write an episode description (max 200 words) optimized for podcast
search. Include 3-5 relevant keywords naturally.
3. Create social media posts:
- X/Twitter: 3 tweets — one pull quote, one key insight, one question
to spark discussion. Each under 280 chars.
- LinkedIn: 1 post, professional tone, 100-150 words.
- Instagram caption: 1 post with emoji, casual tone, include relevant hashtags.
4. Extract a "highlights" list — the 3 most interesting/surprising moments
with timestamps.
Save everything to ~/podcast/episodes/[episode-number]/publish/
- Optional: set up daily monitoring of competitor podcast RSS feeds — fill in the feed URLs; notifications arrive via Telegram.
Monitor these podcast RSS feeds daily:
- [feed URL 1]
- [feed URL 2]
When a new episode drops that covers a topic relevant to my podcast,
send me a Telegram message with:
- Episode title and link
- One-sentence summary
- Whether this is something I should respond to or cover from my angle
For local transcript generation, you can use Whisper (OpenAI). Useful resources: Podcast RSS Feed Spec and Spotify for Podcasters.
Source: hesamsheikh/awesome-openclaw-usecases and its contributors. License: MIT.
Copyright (c) 2026
Translation and editorial adaptation by ClawCases.
>Resources
>Related Use Cases
AI-Powered Earnings Tracker
Stay on top of earnings season across dozens of tech companies without manually checking multiple sources and remembering report dates for every AI/tech earnings report.
AI Video Editing via Chat
Describe your edit to an agent to trim videos, add music, subtitles, or crops — no timeline or GUI. Install two skills, then editing is done by conversation.
Voice Assistant Workflows
Set up voice input and output for the Hermes agent so you can talk to it through a terminal microphone loop, Telegram or Discord voice notes, or a Discord voice channel and get spoken replies — all with the same agent, memory, and skills.