
What happened
Developer tool tiara offers check, sync and verify commands; it resolves only deterministic conflicts, such as both sides adding different functions or JSON items, and passes the rest to Claude Code or Codex for review.
Why it matters
Developers running several AI agents on one repository can avoid repeated merge-and-fix cycles because tiara handles only easily predictable conflicts, leaving genuinely ambiguous edits for a human or AI to decide.
Summaries like this, in your inbox every morning.
The tool emerged from a developer's experiment inspired by Anthropic's October 13, 2026 article "Redesigning source control at Anthropic," which argued that source-control burdens like rebasing and conflict handling should be carried by tooling rather than developers. The developer tested that idea at a small scale, finding that Git itself did not choke at a few AI agents producing a dozen or so pull requests per day, so replacing Git was unnecessary. Instead, the effective parts were predicting conflicts in advance, mechanically resolving what can be decided, and passing only the remainder to AI.
The design reflects that split: tiara uses git merge-tree for prediction to match GitHub's merge behavior, resolves only well-defined patterns such as insert-both additions and structured JSON/YAML changes, and otherwise emits a handoff with commit messages so an AI can see what each side intended. The developer describes tiara as a foundation rather than a finished product, version 0.1, meant to be adapted per team because repository layout, branch names, generated files, and test methods vary.
Pick your industry and the AI tools you use, and get news related to your work every day.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.
Deepseek shipped Deepseek-V4.1-Flash-1017 via its website, app, and API, and published the weights on Hugging…

On John Collison's Cheeky Pint podcast, Zuckerberg and Wang said a 44-minute call and a memo led to Meta Super…

The developer switched ALPHA FORGE's judge model to the brand-new Opus 5.5 on Saturday, Sept 26, and it return…

Across 100 chances to answer questions absent from a six-line sample document, Claude Code on claude -p versio…

An AI running on Claude Code described its parts: a Codex watchdog checking every 1 minute, ChatGPT as design…

In the developer's test on Claude Code 2.1.302, deny rules stopped 9 of 11 risky commands, but /bin/rm secrets…
