This guide touches on how to funnel news, curate, and distribute if needed. This is an automation, and such setups have been around for a long time. AI can help a lot here but is not required. Rather, this is a good example of showing the difference between automation and AI, as some people confuse the two.
Get the News
Most commonly this is done with RSS feeds that sites supply, but you can find and create your own non-official ones. There are various tools that you can use to process the feeds such as Feedly, NewsBlur, and Inoreader. Some can be self-hosted, as I do, like FreshRSS. No matter how you do it, having a reader makes it easier to organize your news feeds.
How AI can help
AI can absolutely be used to make and filter feeds in often complex ways. For example, filtering news from multiple feeds into categories, summarizing, and putting it to a Google Sheet or Doc. Feedly also has AI filters on higher plans and full AI intelligence for enterprise ones.
Automate Posting
Now you want to automate the posting or sharing of news. If you want curation or categorization, as I do on Reddit, you have to do additional setup. Having an automation platform is helpful. IFTTT, Make, and Zapier are common solutions. More complex ones include n8n, which you can self-host locally, on a VPS, or in the cloud. That is a wide subject, so it will be handled in a different guide.
How AI can help
AI can help you set up workflows directly or with help. n8n-as-code is an example of such augmentation. You can also use AI directly inside the workflows, as mentioned in the AI section above.
Tip: This is often a good place for refactoring. While you can have individual flows for various sites and categories, having a unified system increases efficiency and reduces latency.
Share to Platforms
Let’s assume your intent is to share this news. You can tie into various social media platforms or your site via API and webhooks. You can choose to handle each platform separately, or automate from one to the others. How you go depends on your setup and intent. Chaining nodes means propagation delay due to polling, but you may even want to add intrinsic delay if you don’t want tons posted at once.
How AI can help
AI can be used for filtering or in other interesting ways for this flow. You may want to build a vector database of certain material. This might also be a good way to build sources for other projects, such as NotebookLM. Lastly, AI can be used in cases where an API is lacking to either make your own API, build a webhook browser extension, or help with whatever you need to make things work.
Tip: Be very careful about how you set up what information is passed. Not all sites can use all data, and you may have to do custom work for some. For example, Reddit flair is not always passed. You can work around this with TypeScript filtering and, if you own the subreddit, automod rules.
Verify Output
This step is more important than you think. The output has to be accurate and has to look good, too. This means tweaking elements of the automated workflow to get it just the way you want. Each platform may require its own tweaking.
How AI can help
AI can be useful as a summarizer if you are just giving people a piece or need an excerpt that links to the full article. This can be used in more advanced ways for your own blogs, for example using AI to write a summary of the article or news in your own voice for your WordPress blog. AI can also be useful for engagement purposes.
Tip: Sharing is often about engagement. Setting things up to make it easier to see and reply to audience comments is ideal. Organizing content consistently over time also makes it easier to build your own database of up-to-date information on your select interests.
Toolkit Reference
Below are the plugins, extensions, and MCP servers used across the steps in this guide.
Community Plugins
- Everything Claude Code
- Agent workflow optimization, skill system, and development patterns
- Codex Integration
- Code review via OpenAI Codex CLI
- Gemini CLI Integration
- Run Gemini directly from the terminal for filtering and summarization tasks
Official Plugins
These are built-in to Claude Code CLI and can be installed directly from the plugin menu.
- superpowers
- Planning and brainstorming for multi-stage automation pipelines
- context7
- Documentation lookup for n8n, RSS specs, and platform APIs
- firecrawl
- Web scraping for custom feeds and source discovery
- serena
- Semantic code analysis for filter and webhook scripts
- code-review
- Automated code review for automation scripts
- code-simplifier
- Cleanup and refactoring for unified workflow code
MCP Servers
- brave-search
- Web search for finding sources, tools, and prior art
- tavily
- AI-oriented search for research and content discovery
Integrations
- GitHub
- Source code, open-source automation projects, and self-hosted feed tools