smart-logging

Import it once. All your console.* logs become async-aware, pretty Pino logs with timestamps and file names—no other changes needed.

Ideal when you just want to install a logger and let it handle every console.log in the project automatically.

Install

npm install smart-logging
# or
yarn add smart-logging
# or
pnpm add smart-logging

Why smart-logging?

| Before (plain console.log) | After (smart-logging) | |---------------------------------------------------------------|----------------------------------------------------------------------------------------| | Synchronous console writes can block during heavy logging | Console calls are deferred, keeping the event loop responsive | | No timestamp, file, or line information | Every message includes ISO timestamp plus the originating file and line | | Hard to trace errors—stack traces get lost in noisy output | Pretty Pino formatting preserves stacks and structures them cleanly | | Integrating a logger means rewriting every console.* call | Import once; all existing console.* calls are patched automatically | | Adding request/user metadata requires manual plumbing | addTag lets you annotate logs with any key/value scope (requestId, userId, etc.) | | Existing projects without logging need major refactors | Drop in the package, import it in the root file, and you’re done |

1. Enable automatic logging

Do this once in the entry file of your app (the place where everything starts). After this, every file can keep using plain console.log, console.error, etc.—they are all patched automatically.

ESM entry (e.g. index.mjs, app.js in "type": "module")

import 'smart-logging';

console.log('Server booting');   // timestamp + file:line
console.error(new Error('Oops')); // pretty stack trace

Output:

2025-11-08 19:03:12.345  src/server.js:4 - Server booting
2025-11-08 19:03:12.352  src/server.js:5 - Oops
    Error: Oops
        at ...

CommonJS entry (e.g. index.js without "type": "module")

require('smart-logging');

console.info('Ready to roll');
console.warn('Using CommonJS');

Output:

2025-11-08 19:05:01.410  src/index.js:4 - Ready to roll
2025-11-08 19:05:01.411  src/index.js:5 - Using CommonJS

Format: [time] file:line - message

That’s it. No manual logger instances, no extra function calls. Any module that uses console automatically inherits the new behaviour.

2. Optional tags (request IDs, users, feature flags…)

If you want to attach metadata to every log within a request, call addTag. You can use any key name (requestId, userId, featureFlag, ip, etc.). Values must be strings, numbers, booleans, or null—functions, objects, or arrays are rejected to keep logs safe and predictable.

import 'smart-logging';
import { addTag, clearTags } from 'smart-logging';

addTag({ requestId: 'req-123', userId: 'user-99' });
console.log('Started processing');

addTag({ featureFlag: 'beta-payments' }); // merges with existing tags
console.warn('Customer in beta flow');

clearTags(); // reset when the request finishes

Logs emitted inside that async call chain will include requestId, userId, and featureFlag automatically:

2025-11-08 19:03:45.112  requestId: req-123 userId: user-99 featureFlag: beta-payments  api/orders.js:27 - Customer in beta flow

What smart-logging does for you

• 🪄 Monkey-patches console.log, info, warn, error, and debug
• 🕒 Adds ISO timestamps and prettified output via pino-pretty
• 📁 Prints the relative file name and line number that produced the log
• 🧵 Uses AsyncLocalStorage to carry tags across async boundaries
• 🚨 Hooks uncaughtException, unhandledRejection, SIGINT, and process exit to log cleanly
• 🧰 Works in both ESM and CommonJS projects out of the box

Troubleshooting

• Publishing: bump the version before each npm publish.
• Tags not showing: ensure addTag runs before the first log in the async path.
• Want custom formatting? Fork and tweak support/core/logger-core.cjs.

License

ISC