Conventional Commits aren't just for humans anymore. In the age of AI agents and automated pipelines, a structured commit history is the single most important factor in maintaining a clean project narrative.

The Core Specification

The specification is simple: <type>(<scope>): <description>. But why does it matter for GitLog AI?

Human Scenario: Team Standards

Maintain sanity in a team of 50+ developers.

feat(auth): add google oauth provider

fix(ui): repair overlapping sidebar on mobile

docs: update readme with deployment steps

AI Agent Scenario: Automated Versioning

GitLog AI detects feat as a minor bump and fix as a patch bump automatically.

"I noticed you have 3 features and 1 breaking change since v1.0.0. I suggest bumping to v2.0.0 and I have drafted the announcement for you." — GitLog AI Bot

Self-Hosting Guide (CLI)

For teams who want to run GitLog locally or in a private cloud, we provide a CLI wrapper.

# Install the GitLog CLI

npm install -g gitlog-ai

# Generate local changelog

gitlog generate --from v1.0 --to HEAD --format keepachangelog

# Push metadata to MCP

gitlog sync --mcp-token YOUR_TOKEN_HERE

Standardize Your History

Stop guessing what went into your releases. Let GitLog AI enforce the standard.

Get Started