Roadmap

This page tracks what the JuhJuh team is building and where the platform is headed. Features move through three stages: Shipped, In Progress, and Planned. We update this page quarterly.

Want to request a feature? Use the ideas board inside JuhJuh to submit and vote on suggestions.

Recently shipped¶

These features are live and available to all organizations.

Infrastructure as Code¶

Define your entire stack in a single juhjuh.yml file or a .juhjuh/ directory. The JuhJuh File spec supports:

• 6 service types. Web, worker, cron, proxy, daemon, and job.
• Variables. ${var.name} syntax with type coercion for strings, integers, and booleans.
• Templates. Reusable service definitions with recursive inheritance chains.
• Environment overrides. Inheritance between environments, so staging can extend production config with targeted changes. Disable specific services per environment.
• Multi-file configs. A .juhjuh/ directory with numbered YAML files that merge alphabetically using deep merge semantics.
• Deployment references. Wire deployments together with ${deployment.name.host}, ${deployment.name.port.alias}, and ${deployment.name.endpoint.alias}.
• Resource definitions. Databases, caches, queues, and search engines with auto-injected environment variables when services link to them.
• Firewall rules. Per-service and global rules with port, protocol, access level, and source range controls.
• Networks. Organization-scoped networks connecting services across VMs.
• Lifecycle hooks. Pre-deploy and post-deploy commands for migrations, cleanup, and health check grace periods.
• Compute specs. CPU and memory limits per service with automatic reservation estimation.

CLI¶

The JuhJuh CLI gives you full control from your terminal:

• juhjuh init scaffolds a new config file with interactive prompts or --non-interactive mode for CI pipelines.
• juhjuh validate runs 8-phase validation: syntax, variable definitions, variable resolution, schema, vault config, cross-references, environment group references, and online vault verification.
• juhjuh apply provisions VMs, imports vault secrets, and deploys in dependency order. Use --dry-run to preview, --skip-deploy for VMs and vaults only, or --wait to block until completion.
• juhjuh deploy rollback rolls back to the last successful deployment with an automatic version bump.
• juhjuh vault diff shows drift between your current vault and the last deployment snapshot.
• juhjuh vault import and juhjuh vault export move secrets between your config file and the server.
• juhjuh org commands manage your default organization context.
• juhjuh destroy tears down deployments and VMs in reverse dependency order.

Vault management¶

Vault stores secrets per VM with versioning and snapshot tracking. At deploy time, JuhJuh captures an immutable snapshot of your vault entries. After your next deploy, vault diff shows exactly what changed. Import secrets from your config file, export as .env, and set entries individually or in bulk. Secrets are encrypted at rest and injected into your application as environment variables at deploy time.

Vault entries support two variable types (text and file) and four source tags (manual, auto, imported, restored) so you can trace how each secret was created.

Deployment dependencies¶

Deployments can declare dependencies on other deployments with three conditions: healthy, started, or completed. JuhJuh resolves the dependency graph using topological sorting and deploys each level in parallel. Circular dependencies are caught during validation, not at deploy time.

Deployments also support exposed port aliases, scoped resources and services, registry authentication, and file syncing.

Agent pipelines and marketplace¶

Agent Pipelines chain multiple agents into multi-step workflows. The Agent Marketplace lets organizations share and install pre-built agents across teams. Marketplace items support 1-5 star ratings so teams can evaluate agents before installing them.

Code review¶

Code Review integrates AI-assisted review into your ticket workflow. Agents analyze pull requests, flag issues, and suggest fixes before your team reviews.

Schedules and notifications¶

Schedules run agents and pipelines on a recurring basis. Notifications keep your team informed through in-app alerts and Slack messages.

Auto-stop for idle environments¶

Development and staging VMs that sit idle waste money. Auto-stop shuts down VMs after a configurable idle period. Background monitoring checks for idle instances and stops them automatically. Production environments can be excluded.

Cost allocation tags¶

Tag VMs by project, team, or environment and group infrastructure spend. Each tag has a name and color, and you can assign multiple tags to any VM. Cost fields track monthly and hourly rates per instance.

In progress¶

These features are actively being built. Expect them in the next 1-2 quarters.

Advanced cost dashboards¶

Per-VM cost breakdowns with hourly rate calculations, monthly projections, and cost alert thresholds. The billing infrastructure tracks monthly costs and hourly rates per instance today. The next step is a dedicated dashboard that surfaces this data with filtering by tag, project, and environment.

Scaling policies¶

Define CPU and memory thresholds that trigger automatic scaling. The JuhJuh File already supports scaling configuration with minimum and maximum replica counts, target CPU and memory utilization percentages, and separate cooldown periods for scale-up and scale-down. Active enforcement of these policies is in progress.

Reserved capacity discounts¶

Commit to 1-year or 3-year terms for predictable workloads and receive discounted rates. The instance model already tracks commitment type, start and end dates, and discount percentages. The billing dashboard will show your effective rate alongside on-demand pricing so you can decide where commitments make sense.

Live deployment log streaming¶

Stream deployment logs in real time from the dashboard. Today, deployment logs are available after completion. Live streaming will show each step as it runs, with the ability to cancel a deployment mid-flight.

Planned¶

These features are on our roadmap but have not entered active development yet.

Multi-region deployments¶

Deploy services across multiple regions with a single config file. JuhJuh will handle network peering, DNS routing, and data replication between regions. Region-specific overrides will work the same way environment overrides do today.

Overlay networking¶

Organization-scoped overlay networks will connect services across VMs without exposing ports publicly. Today, cross-VM communication uses the organization private network with exposed port aliases. Overlay networking will simplify this by giving every service a routable address within the org network.

Drift detection and auto-remediation¶

Detect when running infrastructure drifts from your declared config. JuhJuh will compare the actual state of your VMs, services, and secrets against your JuhJuh File and flag discrepancies. Vault snapshots already capture state at deploy time for diff comparisons. Full drift detection will extend this to services and compute resources. Optionally, trigger automatic re-deployment to bring everything back in line.

Environment promotion workflows¶

Promote a deployment from staging to production with a single command. JuhJuh will compare the two environments, show a diff of what will change, and require explicit approval before applying. Promotion will respect deployment dependencies and vault differences.

Monitoring and observability¶

The JuhJuh File schema includes a monitoring configuration section. The planned implementation will surface service health, resource utilization, and deployment metrics in the dashboard with alerting thresholds.

How features move through the roadmap¶

1. Planned. The team has scoped the feature and committed to building it, but active development has not started.
2. In Progress. Engineers are writing code. The feature may be available behind a feature flag for early testing.
3. Shipped. The feature is live and documented. It appears in the relevant docs page and changelog.

Priorities shift based on customer feedback and operational needs. If something on this page matters to your team, vote for it on the ideas board or talk to us on Slack.

Related¶

• Ideas - Submit and vote on feature requests
• Infrastructure - Overview of JuhJuh's infrastructure systems
• JuhJuh File - Config-as-code specification
• Billing - Plans, usage, and cost management
• Agents - AI agent configuration and execution
• Deployments - Deploy and manage your services