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

@opendikw/web-server

v0.1.5

Published

Shared DIKW BFF runtime and application capability profiles

Readme

@opendikw/web-server

Shared Node 24 BFF runtime for DIKW applications. MIT licensed; application business code, branding, configuration and static output stay in their own repositories. Native SQLite dependencies remain external at runtime.

import { createServer } from "node:http";
import { registerOutboundInstrumentation } from "@opendikw/web-server/instrumentation";
import { createWebRuntime } from "@opendikw/web-server/runtime";

registerOutboundInstrumentation();
const runtime = await createWebRuntime({
  appId: "dikw-mbweb",
  profile: "mbweb",
  cwd: process.cwd(),
  staticDir: "dist",
  env: process.env,
});
const server = createServer((req, res) => void runtime.handler(req, res));
server.listen(4321);
// The application owns listening and signal hooks. Drain runtime resources
// alongside server.close() during shutdown.

createWebRuntime returns an HTTP handler and idempotent async close. createLogger(scope, appId?) provides the existing structured/redacted startup logger for thin application entries without importing private package paths. Since cohort 0.1.5 it records an Error field as its class name plus a known system or transport code (TypeError [ECONNREFUSED]), never its message, on stdout and in OTel log records; any other object value logs as [object]. Log a reason string yourself only where no provider error can reach it, such as startup config. A handler error that escapes a SERVER span is recorded the same way: the status message is Class [CODE] and exception.type the known code, else the class; its message and stack are not exported. Importing the package does not listen, install signal hooks, read application static files or initialize a database. Factory creation validates configuration, the static build and SQLite, then owns auth sessions, agent sessions and Web jobs. Closing rejects new requests, aborts agent/jobs, drains work and closes databases. The ADK schema and existing HTTP/NDJSON shapes are preserved.

Use /vite and createApplicationPlugins({ appId, profile }) in development. Vite 8 is an optional peer, isolated from the production /runtime dependency graph. The profile guard runs before Vite's Core proxy and both sidecars. Development sidecars retain the existing auth-off behavior; use standalone runtime for OIDC validation. Expose an auth-off development server only locally.

workbench preserves existing routes and role rules. mbweb allows the current paper/read/retrieve, upload/ingest/synth, Wisdom, session, conversion and translation workflows. It denies unknown APIs, maintenance operations, task listing/cancellation, proposals and traces. The application fixes the profile; requests cannot select it. isRequestAllowed exposes the same policy for tests.

Both apps can share a Core knowledge base. This does not provide per-user Core content isolation. Each BFF needs its own session directory, secret and OIDC client configuration. Workbench retains dikw_session/dikw_login; MB uses dikw_mbweb_session/dikw_mbweb_login, all host-only. Agent app names and Web job stores are separate. The #204 user job isolation and #205 sliding renewal, role sync and absolute session cap remain shared behavior.

Existing DIKW_WEB_*, DIKW_AGENT_*, DIKW_CORE_URL and DIKW_SERVER_TOKEN variables are loaded from the application's .env.local, overridden by env. Production MB requires OIDC and a stable, non-secret DIKW_WEB_CORE_ID. Authenticated MB /web/auth/me adds public issuer/coreId; it never exposes the server's Core URL or token. Workbench responses retain their old fields. Agent tools always use the server connection in auth mode. MB registers no maintenance tool; external search/fetch tools require their configured keys.

Request logs carry the application ID; telemetry defaults to its service name, with the standard OTEL_SERVICE_NAME override. ADK/OTel providers are process global: deploy each app in its own process, as intended by the BFF topology.

Build explicitly with npm run build:packages before packing. No installation scripts are needed. Published files contain compiled ESM, declarations, this README and the MIT license; test IdP fixtures and application code are excluded. Source provenance: OpenDIKW/dikw-web commit 162e619d6052d43b8fdc77048b23a60a7b2d4b02; history/authors remain there.

Optional page-scoped Agent sessions bind a canonical Markdown page path at creation and persist it in SQLite. Scoped turns allow only health and reading that exact page; they expose no global retrieval, listing or external tools. The message body cannot change the session scope. Unscoped sessions preserve the existing tool set. Page evidence produces the existing source citation event.