I used Claude for coding, chat, and Q&A, and that’s how most people use it. You paste something in and you get something out. It was useful when the context was limited to a single task. It became more difficult to use once my work spread across multiple tools.

The real friction wasn't generating content; it was in managing everything around it. Over time, I realized that Claude can be more useful when used within the tools I already live in and not as the main tool itself.

Instead of using Claude directly and feeding it the content manually, I connected it to the tools I use regularly. Gmail, Google Sheets, Linear, Google Analytics, and Slack are a few of the tools I use almost daily. None of them are directly related to coding, but they contain a large amount of context that Claude can actually work with.

It wasn’t full automation but a layer where these tools could communicate, giving me better context and a clearer understanding of my workflow. I experimented with this with four different workflows, and each gave me a different kind of understanding.

Email that thinks before you do

Your inbox starts making decisions

Due to the nature of my work, emails are something I can’t ignore. I receive everything from billing notices to collaboration pitches, sponsored post inquiries, and PR requests. Some emails need my attention for replies, while others can be handled with similar responses due to their repetitive nature.

With Claude, I automated most of this process and started getting a structured output each time it ran. The flow was simple. I used Zapier as the automation layer; it read that day’s email and custom labels, then passed context to Claude. Claude read it using the preset prompts and wrote the output to a Google Sheet based on my prompts I set.

What made it easier was I didn’t have to feed Claude the context of each email. It already knew what was happening in each conversation. It could distinguish between important emails, daily digests, and spam. Now, I no longer had to go through each email separately. I check the sheet each day and send the responses Claude generated, with or without edits.

Claude wasn't just filtering the emails; it was deciding how I should respond to them.

Your dev stack without writing code

Claude becomes your silent project manager

When working on multiple side projects simultaneously, development isn’t the problem; keeping track of it is. Each project had its own feature roadmap, issues, and ongoing progress. Though I kept everything documented, switching between each project’s documentation was tedious. I wasn’t always aware of which project needed an urgent bug fix and which could wait.

I was already using Linear for keeping track of the development. Every milestone, issue, and bug was documented there. I connected it to Claude using the native connector. I could now directly ask Claude what was open, what was blocked, and what features or issues had been sitting for too long without any progress.

GitHub was already connected to Linear, and Claude read Linear. It could take advantage of the context from GitHub commits and PR history. It summarized issues and updates in plain language. Once a feature was completed or an issue was fixed, I could update the board by telling Claude, without navigating to Linear myself.

I no longer opened Linear to check things; I just asked Claude, and it gave a structured response. Claude works better when context is already structured; Linear’s labeled issues and milestones give it exactly that.

Analytics that explain themselves

Stop staring at dashboards

For small apps and blogs, Google Analytics is often the default. It is free and gives an enormous amount of data, but for small publishers, it can be overwhelming. This issue is not the lack of data; it’s how it is interpreted. I used GA4 for my side projects and blog analytics.

I was already using Claude for research; connecting GA4 to it made sense. But there was no direct way to connect them. So, I connected Google Drive to Claude. For analytics data, I used Google Sheets with the GA4 Reports Builder extension. It fetched stats for the last seven days and the previous seven days, enough context for Claude to spot patterns and changes.

With the broader context, Claude could give me more accurate insights. It could now flag articles where sessions dropped significantly or bounce rates spiked and log them in a separate sheet tab. It could also review the content and suggest possible reasons behind those changes. For example, for a couple of articles, it suggested that the content was still accurate, but the UI was changed in the app being covered, so adding new screenshots could help.

Google Analytics became something I could ask, instead of something I used to stare at.

Work updates without the noise

Stop switching tabs for context

My work management and discussions as a writer were scattered across two tools, Asana and Slack. One is a project management tool used for editorial management, and the other is for communication. Both worked best in their own environment, but the problem was they didn’t know what was happening in the other. So, I had to manually keep track of progress and channel conversations.

While I was exploring Claude’s native connectors, I noticed it supported both Asana and Slack. From there it was straightforward; Claude could now read both Slack workspaces and Asana projects. There was no need for an automation layer or scheduling tools like Zapier. I had to ask Claude, and it acted as a query layer across both tools.

For example, I could ask Claude what articles were due this week and what was currently in progress. Or I could ask to check the Slack XDA workspace’s main channel, and see if I missed anything important. I didn’t have to open any of the tools to keep myself updated with the progress; Claude handled the context for me.

The layer nobody talks about

You open Claude, input a query, and it generates output. This is how most people use Claude. Using Claude as a standalone tool is useful, but it is the smallest version of what a large language model can do. Once you give it access to your inbox, tracker, and workspace, the way you interact with it changes entirely. For me, it solved different sets of problems, but the outcome was the same; it was much more structured than before, and the overhead of manually moving the context between tools was gone. Once Claude sits inside your workflow, you stop managing information and start asking for it.