Launching today
Trace
Turn GitHub commits into launch content
10 followers
Turn GitHub commits into launch content
10 followers
Trace connects to your GitHub and turns real engineering activity into launch content. It finds the stories hidden in commits, PRs, and repo progress, then generates content angles, weekly summaries, and ready-to-post drafts for X, LinkedIn, and launch updates. Built for founders and builders who want to build in public without manually translating every commit into content.
One thing that would make Trace even better is letting me pick a specific repo or timeframe to pull from, since some commits are internal cleanup and not really launch-worthy. Right now I'm guessing it just grabs everything.
@hayriyezeytun Hey Hayriye, good news: both of those are already in there. You connect GitHub and choose which repo to analyze, so it's never pulling from everything at once. And the commit window is selectable (30, 60, or 90 days) so you can focus on a specific timeframe. The cleanup commit problem is handled differently though: Trace runs a classifier on each commit and only surfaces ones that clear a story threshold, so "fix typo" and "bump deps" get correctly filtered out before generation. Would love to hear what you think once you try it.
Pulled my repo in and it actually surfaced angles I wouldn't have thought to post about, like tying a small refactor to a bigger roadmap shift. Surprised how on-tone the drafts felt for both X and LinkedIn without me babysitting them.
@nimet3paw The refactor-to-roadmap connection is exactly what the Decision pattern is supposed to catch, those commits where the code change is small but the strategic shift underneath it is actually the story. Really glad it surfaced that one without you having to dig for it. Which format did it draft, X or LinkedIn?
Hooked up my GitHub and it actually pulled a real narrative out of a messy sprint I’d never bother writing about. The drafts feel like a starting point, not generic AI slop, which was a nice surprise.
@nilgnbaayauzoi "Starting point, not generic AI slop" is honestly the bar I was trying to hit, so this means a lot. Messy sprints are where the real stories live, the clean ones usually aren't worth posting. Curious what pattern it detected on that sprint, Struggle or something else?
pulled a quick test with a repo and the weekly summary actually sounded like me instead of a template, which was a nice surprise
@erginel28238 "Sounded like me instead of a template" is the hardest thing to get right with AI-generated drafts, and the fact that it landed that way on a quick first test is a good sign. What repo did you test it on, side project or something with more history?