
What happened
A developer released Sib, a command-line LLM client that stores conversations as Git commits instead of using a traditional database like SQLite. Each user prompt and AI response becomes a commit in a Git repository.
Why it matters
By using Git as the storage backend, Sib gives users built-in version control (branching, editing, rewinding), automatic backup, and sharing capabilities without extra code—features that typically require separate implementations in other LLM clients. The approach treats conversations as code-like artifacts with full history and collaboration tools.
What to watch
The project is in early stages and the codebase is not yet stable for large feature work; the creator invites contributions to shape the core API. Installation requires bash ≥ 3.2, Git, jq, curl, awk, and coreutils, and works with any LLM endpoint (defaults to OpenAI).
Summaries like this, in your inbox every morning.
Sib reimagines LLM conversation storage by treating conversations as version-controlled artifacts rather than database records. Instead of building separate features for undo, branching, backup, and sharing—as most LLM clients do—Sib achieves all of them through Git's native capabilities. A conversation is a linear chain of commits, each containing a JSON object with role (user or assistant), content, and metadata. When you ask a prompt, Sib reads the entire chain from HEAD, sends it to the LLM endpoint, and records both the user turn and assistant response as two new commits. This design naturally supports forking (by choosing a parent commit other than HEAD), editing (by creating a new branch from an earlier point), and sharing (by pushing to any Git remote).
The creator built Sib around a deliberate philosophy: rather than shipping each capability as a named feature, let users control context the way Git does. The `-p` flag switches the parent commit, the `-r` flag skips stdin to move HEAD without an API call, and the `-n` flag starts a fresh chain—simple flags that compose into powerful workflows. Named saves (e.g., `sib save eva-3.0-go-off`) let users label important conversation points without memorizing commit hashes. The project is explicitly in early stages; the codebase is not yet stable for large feature work, but the creator invites involvement in shaping the core API.
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.
Cammarata reinterprets InceptionV1's cat-car neuron as a V-shape detector, not two unrelated concepts

Google Deepmind's Gemini 4 Argon scored 97.5 percent on ARC-AGI-3 at a cost of $4,800, second only to GPT-6 As…

Microsoft's Surface Laptop Ultra, using Nvidia's RTX Spark platform, starts at $2,599 (up to $5,899)

Meta's new commercial terms, effective October 30th, appear to let it crawl and train AI on content from websi…

Vercel data published Thursday shows TypeSafe AI's Jev, a lightweight decision model for choosing among predet…

Reflection AI introduced Beam, an open-weight model for agentic coding and long-horizon tasks, with 501 billio…
