Enhanced Developer Productivity
Scope. Canonical home for the developer-facing surface — the application catalogue, scaffolding, the UIMeta-driven configuration UI, and the quantified business case for self-service infrastructure. The architectural intent of the IDP is in practices/platform_engineering.md; the agent-driven enablement is in outcomes/education_enablement.md.
What this repo uniquely brings to developer productivity
1. Application catalogue (canonical)
A growing library of pre-built application modules ships with the repo. Each exists in CloudRun and GKE flavours with a shared Common module:
| Category | Modules |
|---|---|
| CMS / web | Wordpress_*, Ghost_*, Wikijs_*, Strapi_*, Directus_* |
| ERP / business | Odoo_*, Cyclos_* |
| Education | Moodle_* |
| Healthcare | OpenEMR_* |
| Legal | OpenClaw_* |
| Workflow / automation | N8N_*, Activepieces_*, Kestra_*, NodeRED_* |
| AI / LLM | Ollama_*, Flowise_*, RAGFlow_*, N8N_AI_* (see capabilities/ai.md) |
| Search | Elasticsearch_GKE |
| Frameworks | Django_* |
| Reference | Sample_* |
A developer who needs Django on GCP doesn't write Terraform — they apply modules/Django_CloudRun (or _GKE) and supply tfvars.
2. Self-service scaffolding (canonical)
scripts/create_modules.sh— generates a new CloudRun + GKE + Common triple in one command, with all wiring, file structure, and variable mirroring already correct.- Reference modules —
modules/Sample_CloudRun,modules/Sample_GKEare deployable starting points.
3. UI-driven configuration (canonical)
UIMeta tags on every Application Module variable ({{UIMeta group=N order=M}}) drive a generated configuration UI. Variables auto-organise into groups (0–22 for Cloud Run, 0–21 for GKE) with deterministic ordering. Tooling: update_uimeta.py, VARIABLE_GROUPING_RECOMMENDATIONS.md, .agent/VARIABLE_GROUPING_UPDATE.md. Convention details canonical in practices/platform_engineering.md §3.
4. Quantified productivity gains (canonical)
Per BUSINESS_CASE.md:
| Metric | Manual | This repo | Improvement |
|---|---|---|---|
| Setup time per app | 3–5 days | <2 hours | ~95% faster |
| Cost per setup | $3,200 | $200 | $3,000 saved |
| Maintenance for 10-app fleet | 40 h, $4,000 | 2 h, $200 | 95% reduction |
Full quantification: BUSINESS_CASE.md, IAC_AUTOMATION_BUSINESS_CASE.md. Cost-tier specifics in practices/finops.md.
5. Convention-over-configuration
Every Application Module follows the same shape (5–6 files, mirrored variables, standard wiring) so a developer who learns one learns them all. Convention details in practices/platform_engineering.md §3.
6. Single-flag opinionated defaults
The Foundation Modules pre-integrate everything an app needs — enable_iap = true, enable_cdn = true, enable_binary_authorization = true, enable_vpc_sc = true, enable_pod_disruption_budget = true. Each flag wires in a substantial cross-cutting capability covered in its canonical topic (practices/devsecops.md, capabilities/networking.md, etc.).
7. Fast feedback loop
tofu fmt/tofu validatefor static checks.- Cloud Build pipelines for one-trigger deploys (canonical in practices/cicd.md).
- Integration tests in
tests/for regression coverage.
Cross-references
- practices/platform_engineering.md — IDP architecture, conventions, golden paths (the engineering view of self-service)
- outcomes/education_enablement.md — agent workflows, skill guides, certification material (the learning view)
- practices/cicd.md — pipeline that delivers the developer experience
- capabilities/ai.md — AI subset of the catalogue
- capabilities/multitenancy_saas.md — catalogue as marketplace surface
- outcomes/modernisation.md — catalogue as modernisation target