100+ coding prompts top engineers use to ship 5X faster
Claude Code, Codex, and Cursor are on every engineer's stack. Most still treat them like a search bar. Top engineers work from a system, these 100+ prompts are that system. Sign up for The Code and get the prompts free, plus a 5-minute daily newsletter to keep sharpening your edge.
Most founder stories are told from the finish line. This one is told from the starting blocks, with every number still on the table.
Josue Medina is a software developer from Honduras who has been writing code since 2012. In October he built SkoolRank, a free, member-voted ranking of Skool communities, in about two days. It now lists roughly 5,100 communities in six languages.
Starter Story's page for the project shows $1,000 a month, but Josue himself says revenue so far is $0 and no one has bought anything yet. That gap is the most useful part of the story, because it shows exactly where the hard work starts. Here is what he built and what the first two weeks taught him.
A Gap Inside Someone Else's Platform
Skool hosts thousands of paid and free creator communities, and in 2026 it seemed like every creator was launching one. Skool Games turned growing a community into a competition. Yet as a member, Josue had no good way to tell which communities were actually worth joining.
Skool's own Discovery page sorts by its own signals. The "best Skool communities" lists he found were static blog posts, often written by affiliates. What was missing was a simple signal of which communities members genuinely love.
The format came from outbid.lol, a one-page site where people compete for the top spot. Skool already had natural rivals: thousands of owners in the same niches, each with members willing to click a button for their community. So the plan was one clean page, a ranking per category and language, a free daily vote, and a monthly reset so there is always a new race.
Two Rules Before the First Line of Code
Josue admits he did very little validation. His bet was that building a first version in a couple of days and showing it to real people was cheaper than surveying anyone.
He did set two rules from day one. Paid points must never look like votes. And SkoolRank must stay independent from Skool and respectful of it, which means no Skool logo or branding, only data from each community's public About page, and removal of any listing on request. For a ranking site, those rules are the product: if people stop trusting the order, nothing else matters.
Built in Two Days, Filled in Two Weeks
He wrote the first line of code on October 17, 2026, and opened the site to the public on October 19. The stack is short: Next.js, PostgreSQL, and a small VPS running behind Cloudflare.
The real launch work was content, because an empty ranking is useless. He added communities in batches from their public Skool pages: 986 on launch day, about 1,940 two days later, and about 5,100 by the end of October, each with its own page in its own language. Spanish and English came first, Portuguese and French joined on launch day, and German and Italian followed on October 21 once he saw how many non-English communities existed. About 62% of the listings turned out not to be in English, a breakdown Skool itself does not publish.
The first week, October 19 to 25, brought more than 1,000 visitors and more than 400 free votes. Google Search Console showed 196 clicks and 5,340 impressions between October 20 and 25, almost all from people searching for a specific community by name. Boost purchases: none.
How the Money Is Supposed to Work
The model is deliberately transparent. Voting is always free, and anyone can vote once a day. Every month, the top three communities in each category and language win a crown, a silver, or a bronze medal and enter a hall of fame.
Revenue is meant to come from optional boost points, which start at $5 for 100 points and are paid through PayPal. They are always shown separately from free votes, they are never called votes, and they never guarantee a position. Josue works on the project about four hours a day, alone, with no funding. The first monthly champions were set to be announced on November 1, and that is the first real test of whether owners will pay to compete.
What Each Channel Actually Did
Two weeks is not long, but he tracked every channel honestly, including the ones that did not work.
Search was the early surprise. Google picked up the community pages quickly, and impressions peaked above 1,000 a day on October 23 to 25. Then they fell to about 60 a day, the familiar pattern for a new domain with thousands of pages and no backlinks. His fix is to give every page its own data, such as rank, votes, and alternatives in the same category, instead of repeating Skool's description, and to earn real links.
Paid ads bought clicks, not engagement. A Meta campaign running at $1 to $2 a day spent about $12 for roughly 225 visits from Spanish-speaking countries, around $0.05 each and mostly through Instagram. Those visitors barely voted. He also switched off Audience Network, where most clicks were accidental taps.
X has been quiet so far. Daily posts and replies in the Skool niche have produced very low reach. Alongside that, he wrote ten long guides for Skool owners and members in all six languages, covering topics like how to grow a community, what Skool costs, and Skool alternatives. Optional daily reminders by push notification and email are there to bring one-time voters back the next day.
What He Would Change
The clearest lesson came early: building is the easy part. Two days to build, and weeks so far to get people to care. A few points he draws from it:
Line up distribution before writing code. If he started again, he would recruit a few community owners ready to rally their members on day one, instead of launching into silence.
Owners are the real channel. The people who vote are members sent by the owner of a community they love. Cheap traffic from ads rarely converts into that kind of action.
Google tests new sites, then makes you earn it. Early impressions were real, then faded. Pages that only repeat information found elsewhere do not hold their place, while original data and genuine links do.
Building on someone else's platform takes care. Use only public information, never imply an affiliation, and keep money and votes visibly separate. In a ranking, trust is the whole product.
The First Real Test
The November 1 champions announcement was the first milestone on his list. Every community already has a free website badge that updates with its rank, and Josue plans to reach out to the winning owners personally. He is also doing slow, manual outreach to people who write about Skool. Whether owners start buying boosts will tell him if the model works.
You can follow the experiment on Josue's X profile and see the live rankings at skoolrank.com.
