⌘K

v1.1.0 · Latest Release

Commit with
Intelligence,
Any Language.

The production-grade Git Assistant that understands WHY you coded, not just WHAT you changed. Supports Python, JS, Go, Rust, and more.

Get Started Free View Docs

pip install gitsage-cli

Python· package: gitsage-cli

Conventional Commits

GitSage & Ollama

95%+ Accuracy

Intelligence Simulation

Diff context: auth/middleware.py

diff --git a/src/auth/middleware.py b/src/auth/middleware.py

@@ -42,6 +42,19 @@ class AuthMiddleware:

+ def validate_token_expiry(self, token: str) -> bool:

+ """Check if JWT token is expired."""

+ try:

+ payload = jwt.decode(token, options={"verify_exp": True})

+ exp = payload.get("exp", 0)

+ remaining = exp - time.time()

+ return remaining > TOKEN_REFRESH_THRESHOLD

+ except jwt.ExpiredSignatureError:

+ self._invalidate_session(token)

+ return False

Trusted by developers worldwide

< 2s

Analysis Speed

10+ files with GitSage AI Engine

0.3s

Model Load

Optimized warm-start intelligence

~2MB

Binary Size

Lightweight CLI footprint

95%+

Scope Accuracy

Conventional Commit detection

The Sage Pillars

Intelligence by Design

Every feature is built around a single principle: your Git history should be as intelligent as the code you write.

Deep Intelligence

Beyond simple summaries — GitSage provides Reach & Scope analysis for every change, identifying affected modules and preventing unintended side effects.

Provider Agility

Seamlessly switch between GitSage Cloud and Ollama (Local) for maximum privacy. Your architecture, your rules.

Conventional Excellence

Automatically formats every message to the Conventional Commits specification with 95%+ accuracy in scope identification.

Engine First

A pristine Intelligence Layer that lives independently from the UI — ready for Git hooks, IDEs, CI/CD pipelines, and custom dashboards.

Engine Report

Three-Pillar Intelligence

Unlike standard AI commit generators, every GitSage analysis breaks into three distinct tactical sections.

What Changed

A concise summary of the structural modifications — file paths, line deltas, and semantic changes extracted from the raw diff.

Why it Matters

The logical reasoning behind the change, extracted from code diff context — business justification, bug fixes, and performance gains.

Reach & Scope

Identification of affected modules, classes, and dependencies — preventing unintended ripple effects across your codebase.

Privacy First

Cloud Speed or
Local Privacy.
You choose.

Cloud Mode

Use the GitSage Intelligence Layer for high-performance, deep reasoning. Sub-2s analysis on 10+ modified files.

Stealth Mode (Local)

Run 100% locally with Ollama + Mistral. Your code never leaves your machine.

bash

12345

# Switch to local mode
gitsage config --provider local
# ✅ Ollama connected — mistral:7b available
# 🛡️  Stealth Mode active

Stealth Active

Ollama Ready

// Try the live terminal below

gitsage — 80x24

GitSage v1.0.1 (Stable) — build 0x7A24

·Type 'help' to explore the intelligence engine.

LOCAL_TIME: --:--:--

ENC_STAGED: ACTIVE

PROVIDER: GITSAGE_API

Ready to write intelligent commits?

Install in seconds. Works with any Git repository. No lock-in, no subscriptions.