Serverless Collaboration Guide

Philosophy: Zero Infrastructure, Infinite Scaling. Boring-Gemini uses Git as the universal backend for code, state, and knowledge.

1. Plugin Ecosystem (The "App Store")

In Boring, the "App Store" is just GitHub. You don't need a central server to publish or install extensions.

Install any plugin directly from a Git URL:

boring install https://github.com/boring/security-scanner
# Or short syntax
boring install boring/security-scanner

Bundle your tools, prompts, and workflows into a single .boring-pack file:

Initialize: boring pack init --name my-awesome-pack
Build: cd my-awesome-pack && boring pack build
Share: Upload the .boring-pack file to GitHub Releases, or push the repo directly.

Your agent learns as it works. You can transfer this learning to your teammates.

Export your Vector Database (ChromaDB) to a portable file:

boring brain export --output team-knowledge.boring-brain

Your teammate can import this knowledge. It will merge with their existing brain, not overwrite it.

boring brain import team-knowledge.boring-brain

Collaborate on the same project without conflicts using boring sync. This command handles the complex dance of synchronizing SQLite state via Git.

Export: Boring dumps its internal SQLite state (tasks, milestones) to .boring/sync/state.json.
Git: It commits this JSON file and pulls changes from your team.
Merge: It intelligently merges your team's state into your local SQLite.
Push: It pushes your updates.

Simply run this command frequently:

boring sync

You can also add a message:

boring sync -m "Completed the auth module"

This architecture ensures you own your data, and rely on $0 external infrastructure.