How much does zuzu.codes cost?

The starter track is free — read all lessons and practice for free. Full access to every track (current and future) is $14.99/month. Cancel anytime.

How long does each track take?

Each track is designed as a 30-day challenge — one lesson per day, about 15 minutes each. Go at your own pace, but the structure is built around daily consistency.

What's the lesson format?

Each lesson is a student-teacher dialogue with code examples, followed by a hands-on code challenge in an in-browser editor. You read, you understand, then you write real code.

Do I need prior coding experience?

Our beginner track starts from absolute zero — no prior experience needed. Advanced tracks build on earlier ones, and the platform tells you exactly where to start.

How is zuzu.codes different from freeCodeCamp or Codecademy?

zuzu.codes uses a structured 30-day track format with dialogue-based teaching, an in-browser code editor, and gamification (XP, streaks, progress tracking). The format builds genuine understanding through daily practice.

Threshold alerts — Automation Mastery

Day 18 · ~12 min●

Metrics are useful only if you act on them. The simplest action: when a metric crosses a threshold, alert someone. The simplest someone: yourself.

python

import os

results = ["ok", "err", "ok", "err", "err"]
error_count = sum(1 for r in results if r == "err")

if error_count > 2:
    # Self-send a Gmail alert.
    toolset.execute_action(Action.GMAIL_SEND_EMAIL, {
        "recipient_email": os.environ.get("USER_EMAIL", "you@example.com"),
        "subject": f"alert: {error_count} errors",
        "body": "threshold crossed",
    })
    print("alerted")
else:
    print("under threshold")

With 3 errors > 2, this fires the alert. With 1 error, it skips.

Why self-send via Gmail?

Gmail (or a Calendar event whose summary holds the alert text) is a real notification you'll actually see in your own inbox or calendar. The pattern — "on metric crossing, fire a notification" — works the same regardless of channel. In your own production system, you'd swap the action for whatever notification channel you watch (a chat tool, a paging service, SMS).

Why a threshold instead of alerting on every error?

Alerting on every individual error creates alert fatigue — too many pings, you stop reading them. Thresholds ("more than 2 errors", "latency over 1 second", "error rate above 5%") filter the signal. Alert when the situation is bad enough to warrant attention.

Threshold alerts — fire only when it matters

The pattern

python

# 1. Run work, collect outcomes
results = run_batch()
error_count = sum(1 for r in results if r.failed)

# 2. Compare to threshold
if error_count > THRESHOLD:
    # 3. Notify
    send_alert(f"{error_count} errors in last batch")

Three pieces: collect outcomes, compare to threshold, notify on crossing.

Picking the threshold

A static threshold (> 2, > 100ms, > 5%) is the simplest. More sophisticated:

Percentage thresholds ("alert if error rate > 5%") scale better than counts
Sliding window ("alert if 3+ errors in last 5 minutes") catches sustained badness without firing on isolated blips
Anomaly thresholds ("alert if latency is 3x median") adapt automatically

For a lesson: a static count threshold is plenty. Production systems graduate to sliding-window and anomaly when needed.

Channels

In supported apps for this track, the alert channel options are:

Channel	Action	Use when
Email self-send	`Action.GMAIL_SEND_EMAIL` to your own address	One-off alerts, low frequency
Calendar event	`Action.GOOGLECALENDAR_CREATE_EVENT` with summary in subject	Time-based pings; anything you want to see in a calendar view
Sheets append	`Action.GOOGLESHEETS_SHEET_APPEND_GOOGLE_SHEET_ROW`	Audit trail; alerts you'd want to query later

The chat-tool / paging-service alerts you'd see in a real production setup follow the same pattern — the only thing that changes is the action call.

Avoiding alert fatigue

Threshold high enough that real-bad triggers it but normal noise doesn't
Cooldown / dedup — if the alert just fired 5 minutes ago, don't fire again
Severity levels — info/warn/error so you can filter what gets actively paged vs just logged
Escalation — first alert goes to your inbox; second goes to your phone; third to your team

For this lesson: simple count threshold is enough. The real production layer is week-by-week refinement.

One alert per batch (not per error)

A common antipattern:

python

# wrong — one alert per error
for item in items:
    try:
        process(item)
    except Exception as e:
        send_alert(f"failed: {item}")    # 50 errors = 50 alerts

Better:

python

# right — one alert per batch
failures = []
for item in items:
    try:
        process(item)
    except Exception as e:
        failures.append((item, e))

if len(failures) > 0:
    send_alert(f"{len(failures)} failures in batch")

The summary alert is more useful than 50 individual ones — you immediately see the count and severity.

Day 18 · ~12 min●

Metrics are useful only if you act on them. The simplest action: when a metric crosses a threshold, alert someone. The simplest someone: yourself.

python

import os

results = ["ok", "err", "ok", "err", "err"]
error_count = sum(1 for r in results if r == "err")

if error_count > 2:
    # Self-send a Gmail alert.
    toolset.execute_action(Action.GMAIL_SEND_EMAIL, {
        "recipient_email": os.environ.get("USER_EMAIL", "you@example.com"),
        "subject": f"alert: {error_count} errors",
        "body": "threshold crossed",
    })
    print("alerted")
else:
    print("under threshold")

With 3 errors > 2, this fires the alert. With 1 error, it skips.

Why self-send via Gmail?

Why a threshold instead of alerting on every error?

Threshold alerts — fire only when it matters

The pattern

python

# 1. Run work, collect outcomes
results = run_batch()
error_count = sum(1 for r in results if r.failed)

# 2. Compare to threshold
if error_count > THRESHOLD:
    # 3. Notify
    send_alert(f"{error_count} errors in last batch")

Three pieces: collect outcomes, compare to threshold, notify on crossing.

Picking the threshold

A static threshold (> 2, > 100ms, > 5%) is the simplest. More sophisticated:

Percentage thresholds ("alert if error rate > 5%") scale better than counts
Sliding window ("alert if 3+ errors in last 5 minutes") catches sustained badness without firing on isolated blips
Anomaly thresholds ("alert if latency is 3x median") adapt automatically

For a lesson: a static count threshold is plenty. Production systems graduate to sliding-window and anomaly when needed.

Channels

In supported apps for this track, the alert channel options are:

Channel	Action	Use when
Email self-send	`Action.GMAIL_SEND_EMAIL` to your own address	One-off alerts, low frequency
Calendar event	`Action.GOOGLECALENDAR_CREATE_EVENT` with summary in subject	Time-based pings; anything you want to see in a calendar view
Sheets append	`Action.GOOGLESHEETS_SHEET_APPEND_GOOGLE_SHEET_ROW`	Audit trail; alerts you'd want to query later

The chat-tool / paging-service alerts you'd see in a real production setup follow the same pattern — the only thing that changes is the action call.

Avoiding alert fatigue

Threshold high enough that real-bad triggers it but normal noise doesn't
Cooldown / dedup — if the alert just fired 5 minutes ago, don't fire again
Severity levels — info/warn/error so you can filter what gets actively paged vs just logged
Escalation — first alert goes to your inbox; second goes to your phone; third to your team

For this lesson: simple count threshold is enough. The real production layer is week-by-week refinement.

One alert per batch (not per error)

A common antipattern:

python

# wrong — one alert per error
for item in items:
    try:
        process(item)
    except Exception as e:
        send_alert(f"failed: {item}")    # 50 errors = 50 alerts

Better:

python

# right — one alert per batch
failures = []
for item in items:
    try:
        process(item)
    except Exception as e:
        failures.append((item, e))

if len(failures) > 0:
    send_alert(f"{len(failures)} failures in batch")

The summary alert is more useful than 50 individual ones — you immediately see the count and severity.

Threshold alerts — fire only when it matters

The pattern

Picking the threshold

Channels

Avoiding alert fatigue

One alert per batch (not per error)

Threshold alerts — fire only when it matters

The pattern

Picking the threshold

Channels

Avoiding alert fatigue

One alert per batch (not per error)

Threshold alerts — fire only when it matters

The pattern

Picking the threshold

Channels

Avoiding alert fatigue

One alert per batch (not per error)

Sign up to practice

Threshold alerts — fire only when it matters

The pattern

Picking the threshold

Channels

Avoiding alert fatigue

One alert per batch (not per error)

Sign up to practice