TGraph Backend Generator - Project Roadmap

Vision: Become the definitive code generation toolkit that transforms Prisma schemas into production-ready, enterprise-grade full-stack applications with best-in-class developer experience.

Table of Contents

  1. Current State (v0.0.2)
  2. Core Philosophy & Goals
  3. Roadmap Strategy
  4. Roadmap by Milestone
  5. Feature Backlog
  6. Technical Excellence
  7. Ecosystem & Community
  8. Long-Term Outlook
  9. Success Metrics

Current State (v0.0.2)

✅ What Works Today

  • CLI commands (api, dashboard, dtos, all, init) that generate NestJS + React Admin artifacts from Prisma schemas.
  • Prisma schema parsing with directive support for @tg_form, @tg_format, @tg_upload, and @tg_readonly.
  • Automated generation of Nest DTOs, services, controllers, module updates, and React Admin resources (List/Edit/Create/Show/Studio).
  • Safe regeneration via .tg. naming, tagged sections, and module updaters that preserve custom code.
  • Documentation covering getting started, field directives, architecture, SDK usage, and troubleshooting.

⚠️ Gaps Observed

  • Hard-coded file path assumptions (src/app.module.ts, src/dashboard/src/...) make adoption brittle for non-standard project layouts.
  • CLI prompts block CI/non-interactive workflows; there is no --yes, dry-run, or diff preview.
  • Missing environment preflight: absent schema files, Swagger JSON, or formatting tools cause late failures.
  • Limited end-to-end coverage: CLI smoke tests now validate single-run generation, but idempotency and multi-project scenarios still need automation.
  • Dashboard generation depends on a pre-built Swagger JSON without guided automation or safeguards.

Recently Addressed Feedback (Unreleased)

  • CLI
    • tgraph init --help shows command-specific options (--output, --requireAdmin).
    • Interactive init wizard asks for key paths and defaults.
    • New tgraph static command to list and generate specific static modules.
    • New tgraph types command to generate api.ts from an existing swagger.json.
  • API Generation
    • Prompt before generating static files; interactive selection of which to generate.
    • Controller routes now use config.api.prefix (no hardcoded tg-api).
    • Can generate public controller (no guards) by disabling authentication.enabled or omitting guards.
  • Static Files
    • Added feature-flag guard and audit interceptor to static templates and CLI selection.
  • DTOs
    • Create/Update DTOs now include scalar arrays (e.g., string[]) with correct validators.
  • Dashboard
    • Types generation decoupled from dashboard command; use tgraph types after updating swagger.json.

Open items moved to backlog:

  • Static module commands: richer help per template with usage notes.
  • Services: support relation include in responses.
  • Service methods: add getOneBy<UniqueField> for all unique fields.
  • Export return types from generated client; improve api.ts typings/intellisense.
  • Avoid unused imports in dashboard components (e.g., AutocompleteArrayInput).
  • Revisit enum imports for frontend friendliness.

Core Philosophy & Goals

Design Principles

  1. Convention Over Configuration – Ship sensible defaults, reveal complexity only when needed.
  2. Generate, Don’t Replace – Live alongside handcrafted code with clear boundaries and sentinel comments.
  3. Progressive Disclosure – Let teams scale from basic CRUD to advanced customization at their own pace.
  4. Type Safety First – Preserve end-to-end TypeScript types from Prisma schema to frontend client.
  5. Production Ready – Generated code must meet NestJS and React Admin best practices out of the box.

Strategic Goals

  • Deliver a delightful developer experience that feels native to the NestJS + Prisma ecosystem.
  • Earn trust for production deployments through reliability, documentation, and observability.
  • Enable deep customization without forks via directives, templates, and SDK hooks.
  • Grow an engaged community that contributes directives, recipes, and plugins.
  • Expand to additional frameworks only after the core experience is exceptional.

Roadmap Strategy

We will iterate through small, outcome-driven milestones. Each milestone is independently shippable and focuses on widening real-world usability before adding breadth.

Guiding heuristics:

  • Prioritize stability and automation so the generator is predictable in CI and team workflows.
  • Add capability only when we can cover it with tests, examples, and clear docs.
  • Expose extension points early to avoid forking and keep customizations first-class.

Roadmap by Milestone

Superseded Plan Items

  • Phase 1: Foundation & Stability (v0.1.x - v0.5.x) — Replaced with a lean milestone plan to ship a stable v0.1 faster.
  • Phase 2: Enterprise Features (v0.6.x - v1.0.0) — Folded into a focused production-readiness milestone once stability is proven.
  • Phase 3: Multi-Framework Expansion (v1.1+) — Deferred until the NestJS + React Admin experience is undeniably best-in-class.
  • Phase 4: Cloud Platform & SaaS (v2.0+) — Out of scope for the next 24 months; revisit post v1.0 adoption.

Milestone A – v0.1.0 Launch Readiness (target: 4–6 weeks)

Goals

  • Make generation reliable on real-world NestJS + React Admin projects.
  • Support scripted/CI usage without manual intervention.
  • Detect misconfiguration early with actionable guidance.

Deliverables

  • TOP PRIORITY Add non-interactive mode (--yes flag + config option) covering module creation, overwrites, and dashboard regeneration so CI pipelines can run tgraph safely.
  • TOP PRIORITY Validate configuration before execution (schema validation, friendly errors) and ship tgraph doctor to surface missing files, Prisma issues, or incompatible Node versions upfront.
  • TOP PRIORITY Make key project paths configurable (Nest AppModule location, feature/infrastructure roots, dashboard provider & App entry) with intelligent discovery and clear docs.
  • Add a dry-run/preflight command that reports pending changes, missing modules, and required manual steps without touching the file system.
  • Provide official example repositories (API-only, full stack) and add end-to-end tests that run the CLI twice to guarantee idempotent regeneration.
  • Refresh docs with “happy path checklists” (schema markers, Swagger generation, custom guard integration) and troubleshooting for common failure modes.

Milestone B – v0.2.0 Developer Experience (target: 6–8 weeks after v0.1)

Goals

  • Shorten the iteration loop when evolving Prisma schemas.
  • Give developers confidence before committing generated code.
  • Extend directives based on actual user feedback instead of speculative breadth.

Deliverables

  • Implement --watch / tgraph dev mode that reruns relevant generators on Prisma schema changes with clear status output.
  • Add --diff and --output-dir options to preview generated changes or emit to a sandbox directory before applying.
  • Allow overriding templates via project-local hooks (e.g., templates/nest/service.ts.ejs) with merge-safe defaults.
  • Extend directive support with the most requested controls (@tg_json, @tg_richtext, currency helper) backed by docs and automated tests.
  • Improve dashboard UX by generating reusable shared components (file upload, address inputs) and documenting resource customization patterns.

Milestone C – v0.3.0 Production Readiness (target: 8–10 weeks after v0.2)

Goals

  • Ship the first production deployments with minimal manual stitching.
  • Bake in enterprise requirements without overwhelming first-time users.

Deliverables

  • Generate optional RBAC scaffolding (policy classes, decorators, dashboard access control) that plugs into existing Nest guards.
  • Support relation-heavy workflows: nested create/update for many-to-many, configurable cascade behaviour, and safer handling of self-relations.
  • Enhance REST endpoints with richer query options (range filters, nested sorting, soft-delete awareness) and align the React Admin data provider to consume them.
  • Provide “API surface contracts” (OpenAPI generation, typed client) as part of the CLI so dashboard type generation is one command.
  • Harden upgrade paths with migration notes, breaking-change detection, and automated changelog generation per release.

Milestone D – v0.4.0 Extensibility & Integrations (target: after v0.3)

Goals

  • Let teams extend behaviour without forking the core.
  • Validate the architecture against an additional frontend integration once the core is stable.

Deliverables

  • Introduce a stable plugin API (Node hooks) for custom directives, templates, and post-processing steps.
  • Add a proof-of-concept alternate frontend integration (e.g., React Admin + TanStack Router or a lightweight Next.js admin) to demonstrate extensibility.
  • Expose granular SDK commands for selective regeneration (single model, dashboard only) with automation recipes.
  • Evaluate privacy-conscious telemetry opt-in to understand feature usage and guide prioritisation.

Feature Backlog

  • Schema & Backend
    • Parameterise pagination defaults per model (page size, default sort).
    • Offer soft-delete patterns (deletedAt) with matching filters and dashboard toggles.
    • Add request/response transformation hooks for auditing, event dispatching, and search indexing.
  • Dashboard UX
    • Generate layout presets (cards, split panes, tabbed forms) when schema metadata suggests them.
    • Provide bulk actions (export, inline edit) for eligible resources out of the box.
    • Surface help text from schema documentation comments in generated forms and tables.
  • Developer Workflow
    • Publish a VS Code extension for directive IntelliSense and snippet insertion.
    • Add a CLI analytics command to measure generation time and highlight slow steps.
    • Maintain a compatibility matrix for supported versions of NestJS, Prisma, React Admin, and Node.

Technical Excellence

  • Testing – Snapshot generated artifacts, add nightly end-to-end runs on sample projects, and cover error paths (missing schema, conflicting files).
  • Release Engineering – Automate semantic versioning, ship canary builds per commit, and gate main merges on passing e2e suites.
  • Performance – Benchmark against 100+ model schemas, profile file-system writes, and introduce caching or incremental generation where it matters.
  • Observability – Provide structured logging with log levels and a --debug flag; capture generation metrics for troubleshooting.
  • Quality Gates – Enforce formatting/linting on generated output and require documentation updates for any new configuration surface.

Ecosystem & Community

  • Publish and maintain official example repositories (API-only, full stack, dashboard-lite).
  • Enhance documentation with search, versioning, and “copy” snippets for directives and recipes.
  • Ship a monthly changelog and roadmap update summarising progress, upcoming focus, and community highlights.
  • Curate issues tagged good-first-issue and host quarterly contributor calls or livestreams.

Long-Term Outlook

12–24 Month Horizon

  • Reach v1.0 with multiple documented production deployments and a stable API surface.
  • Evaluate adding GraphQL and alternative admin UI targets only after NestJS + React Admin parity is rock-solid.
  • Introduce enterprise-friendly features (workspace templates, policy packs) based on actual adopter feedback.

Superseded Long-Term Items

  • 5-Year Vision (2030) — We will revisit stretch goals like hosted platforms or conferences once the next two-year plan is delivered.

Success Metrics

  • Developer Metrics
    • Time to first generated API: < 5 minutes on a fresh project.
    • CLI success rate: > 98% (tracked via opt-in telemetry or support reports).
    • Documentation usefulness: > 4/5 satisfaction in surveys or community feedback.
  • Quality Metrics
    • Test coverage: ≥ 90% on critical generators and updaters.
    • Idempotency: zero regressions in nightly double-run e2e suites.
    • Generation time: < 10 seconds for a 100-model schema on reference hardware.
  • Adoption Metrics
    • npm downloads: 10K/month by the time v1.0 ships.
    • GitHub stars: 1K+ with active issue engagement.
    • Community: 500+ members across Discord/Discussions with consistent participation.

Contributing to the Roadmap

We welcome community input:

  • Suggest features via GitHub Discussions tagged roadmap.
  • Upvote priorities with 👍 reactions to help us order work.
  • Contribute code through issues tagged help-wanted and align PRs with active milestones.
  • Share case studies so we can prioritise directives and templates that unlock real projects.

Roadmap governance:

  • Monthly check-ins to review milestone progress and adjust scope.
  • Public updates when priorities shift, with rationale.
  • Transparent criteria for promoting backlog items into upcoming milestones.

Conclusion

TGraph Backend Generator is focused on earning trust through reliability, polish, and extensibility. By delivering a stable v0.1, investing in developer experience, and layering production-ready capabilities over time, we can become the go-to toolkit for Prisma-driven teams.

Last updated: January 2026
Next review: March 2026

For questions or suggestions, open a GitHub Discussion or reach out to the maintainers.

Legacy Idea Bank (Reference)

Prior plan items retained for inspiration. Labels reflect current view of priority/fit within the streamlined roadmap.

  • Keep (Medium) Enhanced field directives pack (@tg_richtext, @tg_currency, @tg_phone, @tg_color, @tg_json, @tg_rating, @tg_markdown) — aligns with Milestone B once stability and config flexibility land.
  • Keep (Medium) Advanced validation set (regex/custom validators, cross-field/conditional validation, custom error messages, unified error formatting) — valuable for enterprise teams; schedule after configuration hardening.
  • Keep (High) Relation experience overhaul (many-to-many polish, nested create/update, cascade controls, bidirectional syncing, polymorphic & self-relation UI) — critical for realistic schemas; maps to Milestone C.
  • Keep (Medium) Developer workflow helpers (watch mode, interactive config wizard, diff preview, migration helpers, conflict resolution, performance tuning) — fold into Milestone B/C selectively; wizard likely optional docs-driven.
  • Keep (Medium) Testing scaffolding (fixtures, API integration suites, component/unit tests, data factories, E2E utilities, coverage reports) — incorporate gradually once core e2e smoke tests exist.
  • Keep (High) Authentication & authorization toolkit (RBAC templates, field/resource-level permissions, multi-tenant patterns, auth strategy samples) — required for broad production adoption (Milestone C).
  • Defer (Medium) Workflow automation extras (audit logging, observability bundles, monitoring dashboards) — useful later; focus on core generation first.
  • Keep (Low) CLI ergonomics (diff preview, dry-run, global --yes, telemetry toggle) — partially absorbed in Milestone A; telemetry remains opt-in future work.
  • Keep (Medium) Template overrides / plugin system (custom directives, template hooks, post-processing) — foundational for extensibility; target Milestone D.
  • Keep (Low) Performance enhancements (generation caching, incremental rebuilds) — monitor after real-world feedback to avoid premature complexity.
  • Defer (Low) AI-assisted features (AI schema suggestions, automated optimizations, intelligent code review) — interesting but distracting before v1.0 stability.
  • Defer (Low) Cloud/SaaS platform (hosted generation, collaboration workspace, real-time editing) — revisit once open-source toolkit is mature and resourced.
  • Keep (Low) Ecosystem ambitions (plugin marketplace, conference, certification program, regional user groups) — aspirational community goals; note for long-term once user base scales.
  • Defer (Low) Enterprise guarantees (SOC 2/ISO compliance, dedicated account management, SLA tiers) — requires sizable adoption; capture in future business planning.
  • Keep (Medium) Example & documentation expansion (recipes, tutorials, architecture deep dives, troubleshooting guides) — continue investing as features ship.
  • Defer (Low) Multi-framework expansion (support for additional backend/frontend stacks) — pause until NestJS + React Admin experience is best-in-class.
  • Drop (Low) Overlapping success metrics (duplicated KPI targets from prior plan) — consolidated into current metrics section for clarity.