Generate a daily podcast briefing from newsletters
Going beyond "organize my folder" - honest results from non-obvious experiments you can replicate. Task: The author attempted to automate a daily podcast briefing from newsletters using Claude Cowork and n8n. The workflow involved fetching newsletter content from Readwise via an n8n MCP server, generating a text briefing, and attempting to create audio. Human involvement was required to set up the n8n bridge and review the output, as Cowork lacks direct network access.; tools: Claude Cowork, n8n. Human role: The original post does not say how a person stayed involved.; limitations: Claude Cowork has limited connectors and no direct network access, requiring workarounds like n8n MCP servers for API calls. The author noted that Cowork struggles with tasks requiring external cloud models like high-quality TTS or speech-to-text. This is an early-stage tool, and the specific podcast automation was deemed insufficient for routine use..
Tools
Published on Agent Work Atlas:
Overview
Going beyond "organize my folder" - honest results from non-obvious experiments you can replicate
Workflow
The author attempted to automate a daily podcast briefing from newsletters using Claude Cowork and n8n. The workflow involved fetching newsletter content from Readwise via an n8n MCP server, generating a text briefing, and attempting to create audio. Human involvement was required to set up the n8n bridge and review the output, as Cowork lacks direct network access.
Reported result
The text briefing was generated successfully with an executive summary. However, the audio generation failed to meet quality standards because Cowork could not access external text-to-speech models due to network isolation, resulting in robotic local TTS output. The author concluded this specific workflow was not viable in Cowork and planned to implement it using Claude Code instead.
Human role
The original post does not say how a person stayed involved.
Limitations
Claude Cowork has limited connectors and no direct network access, requiring workarounds like n8n MCP servers for API calls. The author noted that Cowork struggles with tasks requiring external cloud models like high-quality TTS or speech-to-text. This is an early-stage tool, and the specific podcast automation was deemed insufficient for routine use.
Try it
I want to test a daily briefing workflow using Claude Cowork and n8n. First, set up an n8n workflow that connects to my Readwise API to fetch yesterday's newsletter highlights. Configure this n8n instance as an MCP server accessible by Cowork. In Cowork, prompt it to: 1. Fetch the data from the n8n MCP server. 2. Create a Word document with an executive summary and key headlines. 3. Attempt to generate a 2-3 minute audio briefing. Note: Cowork may fail at audio generation due to network isolation; if it uses local TTS, evaluate if the quality is acceptable. If not, stop and document the failure. Do not publish or send any files until I review the text briefing and confirm the audio status. Replace [YOUR_READWISE_API_KEY] and [YOUR_N8N_URL] with actual credentials. Show me the generated text document first for approval before proceeding to audio attempts.
Source reports are not independently verified. Publication date is the catalog publication date, not the original article date.