Git for Product Metadata

Git for Metadata

Version, deploy, and roll back product metadata with a GitHub-style workflow — starting with MongoDB-backed configuration.

MaGit helps teams safely manage product behavior stored as metadata — workflows, dashboards, forms, templates, feature flags, tenant rules, and AI agent configuration. MongoDB is the first core source, with file-based metadata support planned.

npm i -g @x12i/magit
Copied!
See the Workflow
magit-workflow.sh
MongoDB Metadata
Local / Dev DB
magit commit
MaGit / X12I Remote
Immutable object store
magit release create v2
Deploy to Production
Target: prod-cluster-01
$ magit deploy apply --to prod
✓ Plan verified
✓ Pre-deploy snapshot saved
Applying 42 objects... Done.
Rollback if needed (magit rollback)

The Problem

MongoDB config changes should not move by copy-paste

Many teams store critical product behavior inside MongoDB. That metadata controls how the product works, but changes are often moved manually between dev, staging, and production.

Dev MongoDB Raw documents
Manual Copy
Staging MongoDB Overwritten state
Scripts / UI
Prod MongoDB
No diff No release artifact Hard rollback

The Solution

MaGit gives MongoDB metadata a release workflow

Treat MongoDB-backed configuration like a versioned repository. Save versions, push to remote, create releases, review deployment plans, and roll back safely.

Version history

Know exactly what changed, when it changed, and who changed it in your database.

Diff before deploy

Review structural metadata changes side-by-side before they ever reach production.

Named releases

Promote stable, semantic versions instead of copying raw, unverified database state.

Safe deployment

Plan first, apply second. Generate deployment plans to visualize the execution path.

Rollback

Instantly restore a previous production state using automatically captured checkpoints.

Audit trail

Keep a permanent ledger of commits, releases, deployments, and rollback events.

A familiar Git-style workflow for MongoDB

MaGit feels familiar because the workflow maps directly to Git concepts. You keep MongoDB credentials on your machine or CI.

MaGit snapshots selected metadata, stores immutable objects, manages releases, and applies approved deployments to target environments. X12I storage acts as your remote repository.

  • Credentials stay local or in your CI/CD.
  • Snapshots are converted to immutable objects.
  • Target environments are safely updated via plans.
user@machine: ~/metadata-repo
# 1. Save changes locally
magit commit -m "updated workflow config"
# 2. Sync history to remote
magit push
# 3. Create a named release artifact
magit release create workflow-v2 --commit HEAD --from dev
# 4. Preview changes against production
magit deploy plan --release workflow-v2 --to production
# 5. Apply the approved plan
magit deploy apply --release workflow-v2 --to production --plan <planHash>
# If something goes wrong:
magit rollback apply --deployment <id> --mode exact

Built for product metadata and configuration

MaGit is highly specialized. It is not a generic database tool.

Ideal for Metadata

Use MaGit to version and deploy metadata that controls product behavior — whether it starts in MongoDB today or evolves into structured JSON/Markdown repositories.

workflows dashboards forms templates feature flags app settings catalog config tenant config AI agent config CMS-like data Markdown templates prompt files agent instructions policy docs playbooks JSON config folders metadata packs

Not a backup system

MaGit is not for full database backup, transactional data, or high-volume operational collections. It is for managed metadata that defines product behavior and can be safely versioned, reviewed, released, and restored.

orders payments logs analytics events messages full db backup

Sync history. Deploy releases. Roll back safely.

Object store sync

Commit snapshots, push/pull history, and keep immutable objects in remote storage.

Analogy: Git commits and Git remote.

Pipeline deploy

Create releases, plan deployments, apply to environments, and roll back from checkpoints.

Analogy: CI/CD for metadata.
Commit History Layer
C1
C2
HEAD
push pull clone
Deployment Pipeline Layer
Release
v2.0.0 (from HEAD)
Plan
Diff: +12 -3 ~5
Apply
Deployed to Prod
Rollback

Start from the CLI

Get up and running in minutes.

Install MaGit

Install the CLI globally via npm.

npm install -g @x12i/magit

Connect to X12I Storage

Login to set up your remote object store.

export X12I_MAGIT_TOKEN="..."

magit login \
  --storage-url https://api.x12i.com/magit-storage \
  --token-env X12I_MAGIT_TOKEN \
  --tenant customer-workspace

Initialize Repository

Connect your local database and initialize the metadata repo.

export MONGO_URI="mongodb+srv://..."

magit init \
  --driver mongo \
  --url-env MONGO_URI \
  --db customer_db \
  --name customer-repo \
  --storage-provider x12i-http

Deploy with absolute confidence

After using MaGit, you will say:

  • I saved a version of my configuration.
  • I pushed it to X12I.
  • I can see what changed.
  • I can deploy it to production.
  • I can roll back if production breaks.

For teams that ship product behavior via metadata

Workflow platforms

Version workflow definitions, approvals, routing rules, and automation logic.

Internal tools

Track dashboards, forms, admin settings, and feature configuration safely.

SaaS platforms

Promote tenant configuration, feature flags, templates, and catalog metadata.

AI products

Version agent instructions, tool configuration, memory descriptors, and workflow graphs.

Low-code / no-code

Deploy forms, pages, automations, and customer-specific metadata without breaking live apps.

Designed for production safety

MaGit includes planning, protections, snapshots, and rollback workflows out of the box.

Deploy plans

Preview exact insertions, updates, and deletions before applying.

Protected environments

Require plans, confirmations, and pre-deploy snapshots for prod.

Rollback checkpoints

Capture the exact target state before any deployment modifies it.

Integrity gates

Run validation rules before release, deploy, or promote actions.

Large target safety

Preflight estimates, batching, throttling, and confirmation thresholds.

CI-friendly

Use structured JSON output and progress events for automation pipelines.

X12I Storage

Use X12I like GitHub for your metadata repositories. MaGit repositories are private by default and scoped to your X12I workspace, users, services, and CI tokens.

Your database credentials stay local or in CI. MaGit pushes versioned metadata objects to X12I storage, where repositories, releases, and deployment history are managed centrally.

What X12I Stores
  • • Versioned metadata objects
  • • Repository refs & tags
  • • Release information
  • • Deployment history logs
  • • Compressed metadata bundles (planned)
What X12I DOES NOT Store
  • • Your database credentials
  • • Live transactional data

What we are adding next

Expanding beyond MongoDB and scaling for the future.

Public/template repositories

Reusable metadata packs and starter repositories for fast bootstrapping.

Markdown support

Version prompts, policies, playbooks, templates, and documentation-like metadata.

Folder repositories

Commit structured JSON and Markdown folders without requiring MongoDB.

Compressed uploads

Package metadata into efficient upload bundles before sending to the MaGit API.

Faster large-repo handling

Smarter scans, streaming, delta uploads, bulk operations, and better benchmarks.

Dashboard UI

Hosted metadata release management visible to the whole team.

CLI today. Dashboard next.

The MaGit CLI already supports the core workflow. The upcoming hosted dashboard will make metadata release management visible to the whole team.

app.x12i.com/magit
Coming Soon

Future UI Modules

Repo List Commit History Diff Viewer Release Manager Deployment Logs Rollback Button Environment Status Public Templates Markdown Diff Folder Repos Upload Bundles Large Repo Performance

Frequently Asked Questions

Is MaGit a MongoDB backup tool?
No. MaGit is not designed for full database backup or high-volume transactional data. It is specifically built to version and manage MongoDB-backed metadata and configuration.
What kind of data should I use MaGit for?
Use MaGit for workflows, dashboards, forms, templates, feature flags, tenant settings, catalog config, integration mappings, and AI agent configuration.
Where do MongoDB credentials live?
MongoDB credentials stay entirely on your side. They usually live in your local environment variables during development or in your CI platform's secrets manager during automated deployments. X12I never sees your database credentials.
What does X12I store?
X12I stores versioned metadata objects, repository refs, release information, and deployment history. It acts exactly like the remote origin for your MaGit repositories.
Are MaGit repositories public or private?
MaGit repositories are private by default. They belong to your X12I workspace and are only accessible to authorized users, services, or CI tokens. Public/template repositories are a future roadmap item for reusable metadata packs and examples.
Can I roll back production?
Yes. Before a deployment is applied, MaGit can create a pre-deploy snapshot of the target state. If something goes wrong, you can use the `rollback apply` command to restore that exact checkpoint.
Can this work in CI?
Yes. MaGit supports CLI workflows, JSON output formatting (`--json`), progress events, independent deployment plans, and apply commands that are perfectly suited for automated CI/CD pipelines.
How does MaGit handle large metadata repositories?
MaGit includes preflight estimates, batching, throttling, checkpoints, and progress output. The next performance focus is faster scanning, compressed upload bundles, delta sync, and smarter change detection for large metadata repositories.

Bring Git-style releases to your product metadata

Stop moving production configuration by hand. Commit your metadata, review the diff, create a release, deploy safely, and roll back when needed.

npm i -g @x12i/magit
Copied!
Join X12I Storage Preview