Tool comparison
Backstage vs Stackbilder:
Enterprise Portal vs Lightweight Governance
TL;DR
Backstage is Spotify's open-source developer portal with a rich plugin ecosystem and service catalog. Stackbilder generates threat models, ADRs, and test plans without infrastructure to operate. Different scale, different problem, different team size.
Quick comparison
| Feature | Backstage | Stackbilder |
|---|---|---|
| Setup | Kubernetes + PostgreSQL + 2-3 engineers to operate | Web app — no infrastructure required |
| Cost | $22-30/dev/month managed, or platform team cost self-hosted | $0 free / $29 pro / $19/seat team |
| Primary output | Service catalog, developer portal, TechDocs | Threat models, ADRs, test plans |
| Best team size | 50+ engineers with platform team | 1-30 engineers, any team |
| Governance approach | Catalog existing services retroactively | Generate governance upfront |
| AI app hardening | Not the use case | Core use case |
When to use Backstage
Backstage is right for large engineering organizations
Backstage is the right choice when you have the scale to justify it. That means a platform engineering team with budget, 50+ developers who need a service catalog, and existing services that need to be documented, navigated, and integrated with external systems.
The plugin ecosystem — Kubernetes, PagerDuty, GitHub, Jenkins, Snyk, dozens more — is genuinely valuable at scale. Backstage's TechDocs, service catalog, and developer portal address problems that only manifest at large team sizes. If you have those problems, Backstage is a reasonable answer.
- ✓ 50+ engineer organizations with platform teams
- ✓ Service catalog across a large, diverse estate
- ✓ Developer portal with plugin integrations
- ✓ TechDocs for organization-wide technical documentation
- ✓ Teams with Kubernetes infrastructure already running
Be honest: if you have the platform team budget and the engineering scale, Backstage is the right tool and Stackbilder doesn't replace it.
When to use Stackbilder
Stackbilder for teams that need governance without overhead
Most teams don't have a platform engineering function. They have a handful of developers who need threat models, ADRs, and test plans without standing up a Kubernetes cluster and writing YAML plugin configurations.
Stackbilder generates governance artifacts — threat model, ADRs, test plan — from a description of your app's architecture, in roughly 20ms, with no infrastructure to operate. The output is portable Markdown that lives in your repository.
- → 1-30 engineer teams without platform engineering
- → AI-generated app hardening before production launch
- → Pre-launch governance for new projects
- → Teams that need threat models but not a developer portal
- → Projects where Backstage's operational cost exceeds its value
Where they work together
Stackbilder generates what Backstage surfaces
Some teams using Backstage use Stackbilder to generate the initial governance artifacts for new services. Run Stackbilder when a new service starts: get the threat model, ADRs, and test plan committed to the repository. That repository is then indexed by Backstage's TechDocs, making the governance documentation available in the developer portal.
The workflow: Stackbilder at service creation → governance artifacts in Git → TechDocs indexes the repository → governance is discoverable through Backstage. The formats are compatible because both work with Markdown in Git.
This isn't a required integration — it's an observed pattern for teams that have both. Stackbilder generates upfront governance; Backstage surfaces existing documentation. They address different moments in a service's lifecycle.
Detailed comparison
Five dimensions: operational overhead, governance, team size, cost, AI apps
Operational overhead
Backstage
Self-hosted Backstage requires a Kubernetes cluster, a PostgreSQL database, and 2-3 engineers with bandwidth to maintain it. Managed Backstage options (Roadie, Spotify-hosted) reduce operational burden but add significant per-seat cost. Initial setup typically takes weeks.
Stackbilder
Zero infrastructure to operate. Stackbilder is a web application — sign up, describe your architecture, get governance output. No CLI to install, no cluster to maintain, no YAML configuration files. Governance output is portable Markdown in your repository.
Governance approach
Backstage
Backstage catalogs services that already exist. It's a retroactive documentation and discovery layer — great for making existing infrastructure navigable. It doesn't generate security analysis or architectural documentation for services before they're built.
Stackbilder
Stackbilder generates governance upfront — before the service is built. The threat model, ADRs, and test plan are produced at the start of development, not cataloged after the service is in production. Governance is a starting point, not a retroactive documentation effort.
Target team size
Backstage
Backstage's value scales with engineering organization size. The service catalog becomes useful when there are dozens of services to navigate. The plugin ecosystem adds value when there are enough developers to justify the integration overhead. Below ~50 engineers, the operational cost typically exceeds the value.
Stackbilder
Stackbilder is designed for solo founders through small engineering teams. A single developer building a Cloudflare Workers app gets the same governance output as a 20-person team. There's no minimum team size, no minimum service count, and no platform team required.
Cost structure
Backstage
Self-hosted Backstage carries significant hidden cost: the Kubernetes cluster, the PostgreSQL database, and — most importantly — the engineering time to maintain it. Managed options like Roadie run $22-30+ per developer per month. For 20 developers, that's $5,000-7,000+ per year before infrastructure.
Stackbilder
Free tier: 3 scaffolds per month, full governance output, no credit card. Pro: $29/month for unlimited scaffolds and LLM-polished code generation. Team: $19/seat/month. No infrastructure cost, no maintenance overhead.
AI app hardening
Backstage
Backstage is not designed for AI app hardening. It has no concept of STRIDE threat modeling, no ADR generation for AI-built codebases, and no test plan generation. If your use case is hardening an app built with Lovable, Bolt, or v0 before it reaches production users, Backstage isn't the right tool.
Stackbilder
Hardening AI-generated apps is Stackbilder's core use case. Threat model, ADRs, test plan, and governed scaffold — generated deterministically for apps built with any AI tool. Same input, same output every time. No hallucinations, no inference drift.
Common questions
Is Stackbilder actually a Backstage replacement?
No — and we're not trying to be. Backstage is the right answer for large engineering organizations with platform teams who need a service catalog, developer portal, and plugin ecosystem. Stackbilder is the right answer for smaller teams who need threat models, ADRs, and test plans without the operational overhead of running Backstage. Different tools, different problems, different team sizes.
What does Backstage do that Stackbilder doesn't?
Backstage provides a full developer portal: service catalog, TechDocs, plugin integrations with Kubernetes, PagerDuty, GitHub, Jenkins, and dozens of others. It's a platform for documenting and navigating a large engineering organization's existing services. Stackbilder doesn't do any of that. It generates upfront governance artifacts — threat models, ADRs, test plans — for new projects.
Can I import Stackbilder's ADR output into Backstage TechDocs?
Yes. Stackbilder's ADRs are plain Markdown with consistent structure. TechDocs ingests Markdown from your source repository. Some teams run Stackbilder to generate initial governance artifacts for new services, then commit those artifacts to the repository that TechDocs indexes. The formats are compatible.
We're a 10-person team. Which should we use?
Stackbilder. At 10 engineers, Backstage's operational cost — Kubernetes cluster, PostgreSQL database, 2-3 engineers to maintain it — exceeds the value it provides. Stackbilder gives you threat models, ADRs, and test plans with zero infrastructure. Start with Stackbilder and revisit Backstage when you have a platform team budget to justify it.
We already have Backstage. Can Stackbilder still help us?
Yes. Backstage excels at cataloging existing services. Stackbilder generates upfront governance for new services before they're built. Teams using both typically run Stackbilder when starting a new service — to generate the threat model and ADRs — then surface those artifacts in Backstage's TechDocs once the service is registered in the catalog.
Related pages
Governance without the platform team
Threat models, ADRs, and test plans — generated from your architecture description, no infrastructure required. Free to start.