T3 Code passes 15,000 GitHub stars
T3 Code has around 15.1k stars, 3.3k forks and over 2,000 commits. What those numbers do and don't tell you about whether to depend on it.
The pingdotgg/t3code repository currently sits at roughly:
| Metric | Count |
|---|---|
| Stars | 15.1k |
| Forks | 3.3k |
| Commits | 2,122 |
| Open issues | 408 |
| Pull requests | 429 |
| Licence | MIT |
Numbers move; treat these as a July 2026 snapshot.
What's actually informative here#
Stars are a popularity metric and a weak signal on their own — a project with a well-known creator gets stars for existing. Theo's audience alone explains a good share of 15k.
The more useful figures are 2,122 commits and 3.3k forks.
Two thousand commits on a young project means genuine sustained development, not a launch followed by silence. That's consistent with the nightly release cadence — most days ship a build.
A 1:4.5 fork-to-star ratio is high. Forks correlate with people actually running and modifying code, where stars correlate with people seeing a link. That ratio suggests a real user base rather than a purely observational one.
The 408 open issues#
Worth interpreting rather than reacting to. On a project this young and this fast, a high open-issue count is ambiguous — it can mean neglect, or it can mean lots of people using something early enough to hit rough edges.
The evidence points to the second. Issues in this repository get engaged with, and specific reported problems have been fixed quickly. Some of the ones documented on this site — the initialize timeout, the remote sessionId failure — read as exactly what you'd expect from a tool being pushed hard in configurations its authors don't run.
Also note that a meaningful share of the reported problems originate below T3 Code, in the provider CLIs it depends on. That's a structural consequence of the wrapper architecture, not a code quality signal.
Should you depend on it?#
The numbers don't answer this. The project's own description does:
Very early stage — expect bugs.
Take that at face value. It's an unusually honest thing to put in a README, and it's accurate.
Reasonable to depend on if: you're an individual developer, you can absorb a bad build occasionally, you're comfortable pinning a version when something breaks, and the worktree workflow meaningfully improves your day.
Not yet, if: you need documented stability for a team, you're writing a process around it, or you need commercial support. Alternatives →
The mitigations are cheap either way: pin a version, update your provider CLI alongside it, and remember that everything T3 Code does is git underneath — nothing traps you.
Why MIT matters here#
In a young, consolidating category, the licence is a hedge. Bloop, the company behind Vibe Kanban, announced a shutdown in early 2026; the open-source project continues under community maintenance.
If a commercial tool in this space goes away, you lose the tool. If an open-source one loses its maintainer, the code and the forks remain — and the underlying primitive is git worktree, which isn't going anywhere. With 3.3k forks and an MIT licence, T3 Code has a floor that closed products don't.
That's not a prediction about anyone's viability. It's just the reason licence belongs in the evaluation alongside features.
FAQ#
How popular is T3 Code?#
The repository has roughly 15.1k stars, 3.3k forks and over 2,000 commits as of July 2026. The high fork ratio and commit count suggest active real-world use rather than passive interest.
Is T3 Code actively maintained?#
Yes. Nightly builds ship most days and issues receive engagement. The project is in early development in the 0.0.x version range.
Is T3 Code safe to use for production work?#
The project describes itself as very early stage with expected bugs and breaking changes. It's reasonable for individual development work if you pin versions, but it isn't a settled product.
What licence is T3 Code?#
MIT. You can read, fork and modify it freely, which is a meaningful hedge in a young and consolidating tool category.