Quickstart

This page gets you from pip install to structured production logs in five minutes.

Install

pip install logcore

A logger in two lines

from logcore import get_logger

log = get_logger("myapp")
log.info("Application started")

You get colored human-readable output on stderr by default. For machine-readable JSON:

log = get_logger("myapp", json=True)
log.info("user login", user="alice", role="admin", success=True)

{
  "timestamp": "2026-05-27T10:30:45.123456+00:00",
  "level": "INFO",
  "logger": "myapp",
  "message": "user login",
  "user": "alice",
  "role": "admin",
  "success": true
}

Configuration

Either pass arguments to get_logger:

log = get_logger(
    name="myapp",
    level="INFO",                          # DEBUG, INFO, WARNING, ERROR, CRITICAL
    json=True,
    file="/var/log/myapp.log",
    max_file_size=10 * 1024 * 1024,        # 10 MB
    backup_count=5,
    redact_fields={"password", "secret"},
)

Or set environment variables:

export LOGCORE_LEVEL=DEBUG
export LOGCORE_JSON=true
export LOGCORE_FILE=/var/log/myapp.log
export LOGCORE_REDACT_FIELDS=password,token,secret

See the configuration reference for the full list of options.

Five things you’ll actually use

1. Structured fields

Pass any keyword arguments — they become first-class fields in the log record:

log.info("order placed", order_id=42, amount=99.99, currency="USD")

2. Exception logging with traceback

try:
    risky_operation()
except Exception:
    log.exception("operation failed", operation="risky_operation")
    # Includes the full traceback automatically.

3. Operation timing

with log.time("database_query"):
    result = run_query()
# Emits: "Starting database_query" then "Completed database_query duration_ms=234.56"

Works the same way in async code with async with log.time(...) — auto-detected.

4. Correlation IDs for request tracing

with log.with_correlation_id("req-abc"):
    log.info("processing")
    do_work()
    log.info("done")
# Every record under this block carries cid=req-abc.

See correlation IDs for async, middleware, and cross-task patterns.

5. Sensitive data redaction (automatic)

log.info("user data", username="alice", password="hunter2-very-long")
# Output: ... username=alice password=hu*** ...

Default redacted fields: password, secret, token, key, api_key, auth, etc. Customize via redact_fields=.

What’s next