npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2026 – Pkg Stats / Ryan Hefner

nestjs-ddd-doctor

v0.5.0

Published

Zero-dependency architecture linter for NestJS projects with DDD leanings: thin controllers, pure domain, ports over concrete infrastructure.

Downloads

955

Readme

nestjs-ddd-doctor 🩺

Zero-dependency architecture check-up for NestJS projects with DDD leanings — in the spirit of react-doctor, but for your API's layer boundaries.

Run it, get a check-up from the botsito 🤖, a graded report of where your architecture is leaking, and a one-keystroke handoff to your AI of choice to fix it:

$ npx nestjs-ddd-doctor

        .----.
       ( @  @ )
        \ -- /          nestjs-ddd-doctor
     .--'----'--.       NestJS architecture check-up
    /|    ++    |\
   d |    ++    | b     patient: apps/api/src
     |  .----.  |       profile: pragmatic
     |__|    |__|
        | () |
        '----'

🔴 controller-db — Controller accesses the database directly (move it to a service/repository) (6)
   apps/api/src/whatsapp/bot-config.controller.ts:2
   apps/api/src/whatsapp/whatsapp.controller.ts:18
   ...

🟠 controller-logic — Business logic in controller (dense branching — extract to a service) (1)
   apps/api/src/whatsapp/whatsapp.controller.ts:43

Grade: D   Score: 52/100  (🔴 4 × -10   🟠 2 × -4   🟡 0 × -1)
Needs treatment. Start with the 🔴 findings.

Fix with AI?  [1] Claude Code   [2] Codex   [3] Copy prompt   [enter] skip

No install, no dependencies, one npx. Node ≥ 20. Colored output on TTYs (respects NO_COLOR); plain text when piped. In a monorepo with no ./src it auto-detects the NestJS source dir by looking for *.controller.ts files (e.g. apps/api/src).

Usage

npx nestjs-ddd-doctor [srcDir] [options]

--profile=pragmatic|strict    # default: pragmatic
--ai=claude|codex|clipboard   # skip the menu, hand findings straight off
--ci                          # no prompts; exit 1 on any 🔴 finding
--json                        # machine-readable output (score, grade, findings)
--update-baseline             # freeze current findings; future runs report only new ones

Profiles

pragmatic (default) — for teams that want layered architecture without the full ceremony. Checks the boundaries almost every layered NestJS codebase agrees on: controllers stay thin, the domain stays pure, the application layer talks to ports.

strict — real tactical DDD, the whole thing: controllers orchestrate use cases (not services), repositories implement ports, domain entities live in domain/, persistence models in infrastructure/persistence, and every meaningful feature module has its bounded-context skeleton (domain/ + application/).

npx nestjs-ddd-doctor --profile=strict

Rules

Pragmatic (always on)

| Rule | Sev | What it catches | |---|---|---| | controller-db | 🔴 | ORM imports (drizzle-orm, typeorm, @prisma/client, mongoose, …), @InjectRepository/@InjectModel/DB tokens, or local db/schema imports inside *.controller.ts | | controller-fetch | 🔴 | fetch( / axios. / got( inside a controller — outbound HTTP belongs in infrastructure | | domain-purity | 🔴 | domain/ files importing @nestjs/*, any ORM, or infrastructure/ — the domain stays plain TypeScript | | domain-imports-application | 🔴 | domain/ importing from application/ — dependencies point inward | | controller-logic | 🟠 | More than 5 if/for/switch/while in one controller file — that's orchestration, extract a service | | application-concrete-infra | 🟠 | application/ importing from infrastructure/ — depend on the port, not the class | | handler-io | 🟠 | handlers/ doing direct DB/HTTP instead of using their context's repositories | | no-forward-ref | 🟡 | forwardRef( — circular modules; rethink the dependency direction | | raw-sql-outside-infra | 🟡 | Raw sql` /.query( outside db/, database/, infrastructure/ |

Strict (added with --profile=strict)

| Rule | Sev | What it catches | |---|---|---| | controller-bypasses-use-case | 🟠 | Controller constructor injecting *Service classes — in strict DDD, controllers orchestrate *UseCase/*Query/*Command | | repository-without-port | 🟠 | class FooRepository without implements — program to the abstraction, inject via DI token | | entity-outside-domain | 🟡 | *.entity.ts outside domain/ (ORM persistence models belong in infrastructure/persistence, mapped in the repository) | | missing-bounded-context-layers | 🟡 | Feature module (with a controller and ≥6 files) lacking domain/ or application/ | | fat-service | 🟡 | Service >250 lines with direct DB access — split into use cases + a repository port |

Scoring

Per rule, not per finding: penalty = weight × (1 + log₂(count)), capped (🔴 w10/cap30 · 🟠 w4/cap18 · 🟡 w1/cap8). Score = 100 − Σ penalties (floor 0). Grades: A+ ≥98 · A ≥90 · B ≥75 · C ≥60 · D ≥40 · F.

Why log damping: 200 hits of one rule are usually one systematic decision (e.g. "use cases inject the concrete repo"), not 200 independent sins — linear scoring nukes any large codebase to 0 and stops differentiating. When a single rule drives most of the penalty, the report says so explicitly.

The banner also shows structure recognition (29/41 modules DDD-shaped): the score measures boundary integrity, the structure line credits the attempt — a DDD-shaped codebase with inverted arrows scores worse than it looks, and that's the point.

AI handoff

When there are findings (and you're on a TTY), the doctor offers to hand them off:

  • [1] Claude Code / [2] Codex — writes ddd-doctor-report.md (findings + per-rule fix guidance + ground rules like "don't change behavior, keep tests green") and launches the CLI pointed at it.
  • [3] Copy prompt — same report, straight to your clipboard, paste it into any AI.

Non-interactive: --ai=claude, --ai=codex or --ai=clipboard. CI: --ci disables the menu entirely.

Baseline — adopt it in a legacy codebase

You don't fix 40 findings on day one. Freeze them:

npx nestjs-ddd-doctor --update-baseline   # writes ddd-doctor-baseline.json — commit it

From then on, runs auto-load the baseline and report only new findings — existing debt is tolerated, regressions fail. Counts are tracked per file::rule, so unrelated edits shifting line numbers don't break it. Shrink the baseline as you pay debt down (re-run --update-baseline after fixing).

GitHub Action

Fail PRs that introduce new architecture violations:

# .github/workflows/architecture.yml
name: architecture
on: [pull_request]
jobs:
  ddd-doctor:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-node@v4
        with: { node-version: 20 }
      - uses: lucasscurtoo/nestjs-ddd-doctor@main
        with:
          src-dir: apps/api/src   # optional — auto-detected
          profile: pragmatic      # or strict

Combined with a committed baseline, this becomes a free architecture guard-rail: existing debt passes, new leaks fail the check.

JSON output

npx nestjs-ddd-doctor --json
{
  "src": "apps/api/src",
  "profile": "pragmatic",
  "score": 62,
  "grade": "C",
  "counts": { "high": 3, "med": 1, "low": 2 },
  "baselined": 17,
  "findings": [
    { "rule": "controller-db", "severity": "high", "file": "src/foo.controller.ts", "line": 12 }
  ]
}

Config

./ddd-doctor.config.json:

{
  "profile": "strict",
  "rules": { "controller-logic": "off" },
  "exempt": ["health.controller.ts", "lock.service.ts"]
}

exempt entries are path suffixes; a matching file is skipped by every rule. For a single justified line:

// ddd-doctor-disable-next-line
const rows = await this.db.execute(sql`select 1`);

Philosophy

  • Heuristic, not a compiler. Plain regex over your source. Fast, zero deps, occasionally wrong — that's what the escape hatch is for.
  • Boundaries, not dogma. Pragmatic mode doesn't demand aggregates or CQRS; strict mode is there when you do want the whole tactical toolbox — and the grade tells you honestly how far you are.
  • Convention-driven. Layers are detected by folder name (domain/, application/, infrastructure/, handlers/, db/). If your project doesn't use a folder, its rules simply never fire.

License

MIT