Why I Built 150 Tools Instead of One

Michael · April 2026 · 4 min read

The question I get most is "why 150 tools instead of making one really good one?"

The honest answer is that I started with one. Stockyard was originally an LLM proxy. A single Go binary that sits between your app and OpenAI, tracking costs, caching responses, and handling failover. That was the whole product.

Then I realized the architecture generalizes. A Go binary with embedded SQLite, an HTTP server, a REST API, and a web dashboard. That pattern applies to webhook inboxes, status pages, feature flags, error tracking, analytics, and basically any tool that stores data and shows it in a browser.

The pattern

Every Stockyard tool works the same way. Download a ~12MB binary. Run it. It starts an HTTP server with a dashboard at /ui and a REST API. Data goes in SQLite with WAL mode. No Docker required, no external database, no package manager.

Once I had the template, building each tool was fast. The hard parts are shared: the HTTP server, the SQLite connection pool, the dashboard framework, the license validation, the health checks. What changes per tool is the data model and the domain logic.

This is not a unique insight. Every SaaS tool is fundamentally "a database with a web UI." The difference is that SaaS tools run on someone else's server and charge you monthly for the privilege. Stockyard tools run on your server and store data in a file you own.

The tradeoff

150 tools means none of them are as deep as the dedicated SaaS version. Seismograph is not Sentry. It does not have performance monitoring, session replay, or 50 language-specific SDKs. Paddock is not Statuspage.io. It does not have managed email delivery or an Atlassian integration.

But for a lot of teams, "good enough and self-hosted" beats "perfect and $79/mo." Especially when the self-hosted version installs in 30 seconds and the data never leaves your server.

The bundle bet

The bet is that the bundle is worth more than any individual tool. If you use 6+ tools, $1 first month for all 150 is cheaper than buying them separately. And you get to cancel 6 SaaS subscriptions, which means 6 fewer vendor logins, 6 fewer privacy policies, and 6 fewer surprise price increases.

A typical SaaS stack for a small team: Sentry ($26), Statuspage ($79), LaunchDarkly ($50), Bitly ($35), Mixpanel ($20). That is $210/mo for 5 tools. Stockyard Complete replaces all of them — $1 first month, then $29/mo. The savings compound fast.

Where I am now

Zero paying customers. $0 MRR. 12 free cloud signups. 31 installs in the last 30 days. I publish these numbers publicly at stockyard.dev/open because pretending things are going better than they are is a waste of everyone's time.

The product works. The constraint is distribution. Getting developers to try a tool they have never heard of, from a company with no track record, is the hard part. The site is built, the tracking is wired, the conversion funnel is instrumented. What is missing is people.

Should you try it?

If you are curious, pick one tool that solves a problem you have right now. Install it. Use it for a week. If it works, try a second one. If it does not work, email me at hello@stockyard.dev and tell me what broke.

The whole catalog is at stockyard.dev/tools. Every tool has a free tier with no time limit.