My 2026 Tool Stack (and What I Would Change)

A real, no-affiliate, what's-actually-on-my-Mac list of the 14 tools I use daily, plus the 3 I would change tomorrow if I had to start over.

The rules for this listI have seen about 50 "my 2026 tool stack" posts. Most of them are affiliate-link farms dressed up as recommendations. Most of them list 30+ tools, none of which the author actually uses daily. Most of them are the kind of post that makes you feel like you need to buy 12 new subscriptions to keep up. I do not like those posts. This is not one of them.The rules for this list: 14 tools, all of which I actually use daily, no affiliate links, no tool I would not pay full price for tomorrow. If the tool is on the list, I would pay for it. If I would not pay for it, the tool is not on the list. The list is not aspirational. The list is the actual tools on my Mac right now, in the dock and the menu bar.The 14 tools, in order of how often I open them1. Mac Terminal (built in, $0)I open the terminal probably 80 times a day. Bash, ssh, vim, git, curl, jq, a dozen other things. The terminal is the tool I would give up last. The terminal is the cheapest tool on this list, and the highest leverage. I have a shell script in my dotfiles that sets up aliases for the things I do most often, and the script saves me maybe 30 minutes a day.2. Vim (built in for any Unix, $0)Vim is my editor. I use it for everything: code, markdown, config files, blog posts (I write markdown in vim, convert to HTML with pandoc, drop in the static site folder). Vim is the tool I have used for 15 years, and the tool I will use for another 15. The investment in learning vim is the investment that pays off the longest. The vimrc I have today is not the vimrc I had 5 years ago, and the upgrades have been worth it.3. Raycast ($8/month, the only subscription I would fight to keep)Raycast is the launcher, the clipboard manager, the snippets tool, the calculator, the window manager, the script runner, the quicklinks, and a dozen other things. The $8/month is the only subscription I would actively fight to keep, because the launcher is the part that touches every other tool. The snippets are the part that saves me 30 minutes a day. The window manager is the part that makes multi-monitor setup not painful. The $8/month is the cheapest $8/month I spend.4. 1Password ($3.99/month, family plan)Password manager. I have about 600 passwords. I do not know any of them. 1Password knows all of them. The $3.99/month is for the family plan, which covers my partner and me. The cost is the cost of not having a password manager. The cost of having one is small. The cost of not having one is the cost of every security incident that ever happened to me, multiplied by the chance the next one happens tomorrow. The $3.99/month is the cheapest insurance I buy.5. Fastmail ($5/month, the only email I trust)Email. I have my main email on a self-hosted Postfix + Dovecot setup (the one from [the self-hosting post), but I keep Fastmail as a backup for the addresses that have to be reliable. The $5/month is for the cheapest tier, which is fine for what I use it for. The deliverability is the part that matters, and Fastmail's deliverability is what self-hosting's deliverability is not (yet).6. Obsidian (free for personal use, $50/year for commercial use I do not need)Note-taking. I have about 4,000 notes in Obsidian. The notes are connected, searchable, version-controlled (I sync via Git), and local-first. The local-first is the part that matters. The notes are mine. They will be mine in 10 years. They are not in a cloud that will change its pricing or get acquired and shut down. The free tier is what I use. The $50/year tier is for commercial use, which I do not have.7. Cron (free for the basic tier, $9/month for the notifications I do not need)I am not talking about the cron daemon. I am talking about crontab.guru, which is the cron schedule parser. The free tier of the website is the only thing I use. The $9/month tier is for the hosted cron, which I do not need because I have a real Linux box. The website is one of the few tools I would pay for even if the company shut down the paid tier, because the website is the only place I can paste a cron expression and see what it means.8. Tailscale (free for personal use, $5/month per user for team)VPN. I run Tailscale on every host I care about, and on my phone and laptop. The mesh network is the part that lets me ssh into a Hetzner box from a coffee shop without exposing the box to the public internet. The free tier for personal use is the only thing I use. The $5/month team tier is for work, which I do not currently have through Tailscale.9. Backblaze B2 ($0.05/month, the off-host backup)Object storage for the off-host backup from [the backup script post. The cost is the cost of a few gigabytes of tarballs uploaded nightly. The $0.05/month is the cheapest insurance I buy. The cost of not having the off-host backup is the cost of every disk failure that ever happened to me, multiplied by the chance the next one happens tomorrow. The $0.05/month is the cheapest insurance I buy.10. Hetzner (~$5/month for the cheapest dedicated box)Hosting. I have a dedicated server in Helsinki for $5/month. The server runs the static site, the email server, the cron jobs, the prompt experiments, and a half-dozen other things. Hetzner is the cheapest dedicated hosting I have found that is not awful. The cost is the cost of not having my own box. The cost of not having my own box is the cost of every managed-hosting surprise I have ever had. The $5/month is the cheapest $5/month I spend.11. Notion (free for personal use, I use it for 1 thing)The 1 thing I use Notion for is the post pre-flight checklist (the 4 prep steps from [the 80/20 post). The free tier is what I use. The team tiers are expensive and I do not need them. Notion is the tool I have the most mixed feelings about, because the tool is good for what I use it for and bad for almost everything else. The list is short on purpose.12. Pandoc (free, the conversion tool)Pandoc is the markdown-to-HTML converter. I write the blog posts in markdown, convert with pandoc, post-process with a Python script, drop the result in the static site folder. The pipeline is the one I described in [the flagship post. Pandoc is the part that does the heavy lifting. Pandoc is also the tool I have used the longest on this list, after vim and the terminal.13. Gumroad (free for sellers, takes a cut of sales)Where the [Blog Writing Factory lives. The free tier is what I use. The cut is the cost of the platform. The platform is the cheapest place to sell a $14 digital product, and the cheapest place is the part I am optimizing for. The cut is worth the cost of not running my own checkout.14. Browser (free, the tool I have the most mixed feelings about)Chrome on the Mac, Firefox on the personal Linux box. The browser is the tool I spend the second-most time in, after the terminal. The browser is also the tool I am most likely to quit (e.g. for Arc, for Brave, for something else) and the tool I am least likely to actually quit. The browser is the most-replaced tool on this list, and the least-likely-to-actually-be-replaced tool on this list. The mixed feelings are real.What I would changeIf I had to start over and pick 14 tools from scratch, I would keep 11 of these. The 3 I would change are:1. Notion → Obsidian only. The pre-flight checklist I keep in Notion would live in Obsidian as a note template. The reason it is in Notion is historical (I was using Notion before Obsidian, and never migrated). The reason I would change it is consolidation. One tool for notes is better than two tools for notes. The change is small but real.2. Chrome → Arc or Brave. I have been meaning to switch for about 18 months. The reason I have not is muscle memory. The reason I would change is that the cost of muscle memory is small, and the cost of being on Chrome (Google's data collection) is large. The change is the one I am most likely to actually do in the next 6 months.3. Hetzner → a Raspberry Pi cluster at home. The Hetzner box is the right choice for now. The change I would make is for the long term: a small home cluster for the things that do not need to be on a public host, and a cheap Hetzner box for the things that do. The change is the biggest one of the three, and the one I am least likely to do in the next 12 months, but it is the right end state.What the list does not includeNo social media tools. I do not have a "social media scheduler." I post when I have something to post, which is 2-3 times a week. The cost of a scheduler is the cost of an obligation to post on a schedule, and the obligation is the cost I am not willing to pay. The post is the part that matters, not the schedule.No analytics tool. The site has no analytics (yet). I will add one when the traffic justifies it, but the default is zero analytics, because the cost of an analytics tool is the cost of watching the dashboard, and the watching is the part that does not help me write.No AI tool that I do not actively use. Claude Pro and ChatGPT Plus are both on the list of "I would pay full price for tomorrow." Midjourney is on the "I would not pay full price" list and is therefore not on this list. The list is the actual list, not the aspirational list.The honest partThe list is small on purpose. The list is small because small lists are easier to maintain, easier to switch out of, and easier to be honest about. A 30-tool list is a list of "things I have tried." A 14-tool list is a list of "things I use." The difference is the difference between a museum and a workshop. I want a workshop.The list also changes. I do a quarterly audit of the list, and the audit is the same one I do for AI tools (the [30-minute audit, adapted for non-AI tools). The audit is the part that keeps the list honest. The audit is also the part that determines whether the list of 14 is the list of 14 I would have today, or the list of 14 I would have had a year ago. The list is alive, not static.Want a prompt pack to audit your own tool stack?The Blog Writing Factory includes a "research prompt" that is built for exactly this kind of audit. Feed it your tool list, get back a structured comparison, decide in 20 minutes what to keep, what to kill, what to try. The audit is the system.Buy on Gumroad. $14

Send me the rough edges if you try it. I read every message.