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.

Parsing failures — Ai Foundations

Day 13 · ~11 min●

Models try to follow your format. They mostly succeed. They occasionally don't — "Here's the JSON: {...}" or markdown-fenced code blocks instead of bare JSON. Your code has to handle both.

python

import json
from pydantic_ai import Agent

def classify_with_retry(text):
    prompt = f'Classify "{text}" — return JSON {{"sentiment": "..."}}. Return only valid JSON.'
    for attempt in range(2):
        result = Agent(model).run_sync(prompt)
        try:
            return json.loads(result.output.strip())
        except json.JSONDecodeError:
            prompt = prompt + "\n\nReturn ONLY valid JSON. No markdown, no prose."
    raise ValueError("model failed to return valid JSON after 2 attempts")

A retry that reinforces the format constraint?

Yes — the second attempt's prompt nudges harder. Often that's enough. If it still fails, raise rather than swallow — silent fallback is worse than a clear error.

Parsing failures — `try/except json.JSONDecodeError`

The robust pattern:

python

import json
for attempt in range(2):
    result = Agent(model).run_sync(prompt)
    try:
        parsed = json.loads(result.output.strip())
        break
    except json.JSONDecodeError:
        prompt += "\n\nReturn ONLY valid JSON. No markdown, no prose."
else:
    raise ValueError("model failed to return JSON")

The for/else construct runs the else block when the loop completes without break — clean way to express "exhausted retries".

Common ways the model slips

Slip	Output
Markdown fence	```json\n{...}\n```
Prose preamble	`"Here's the JSON: {...}"`
Trailing commentary	`{...} - this classifies as positive.`
Smart quotes instead of straight	`{"sentiment": "positive"}` (different quote chars — still a parse error)

Mitigations beyond retry

Strict prompt — "Return ONLY valid JSON, no prose, no markdown" — works ~90% of the time.
Strip markdown fences before parsing — defensive: result.output.strip().lstrip('```json').rstrip('```').strip() — helps when the model uses fences but is otherwise valid.
pydantic_ai's typed output — output_type=YourModel handles parsing internally and retries automatically. The cleaner production answer.

When to give up

If 2-3 retries fail, the prompt itself is the problem. Either the task is too ambiguous, or the model is too small. Switch model or simplify the schema before adding more retries.

Day 13 · ~11 min●

Models try to follow your format. They mostly succeed. They occasionally don't — "Here's the JSON: {...}" or markdown-fenced code blocks instead of bare JSON. Your code has to handle both.

python

import json
from pydantic_ai import Agent

def classify_with_retry(text):
    prompt = f'Classify "{text}" — return JSON {{"sentiment": "..."}}. Return only valid JSON.'
    for attempt in range(2):
        result = Agent(model).run_sync(prompt)
        try:
            return json.loads(result.output.strip())
        except json.JSONDecodeError:
            prompt = prompt + "\n\nReturn ONLY valid JSON. No markdown, no prose."
    raise ValueError("model failed to return valid JSON after 2 attempts")

A retry that reinforces the format constraint?

Yes — the second attempt's prompt nudges harder. Often that's enough. If it still fails, raise rather than swallow — silent fallback is worse than a clear error.

Parsing failures — `try/except json.JSONDecodeError`

The robust pattern:

python

import json
for attempt in range(2):
    result = Agent(model).run_sync(prompt)
    try:
        parsed = json.loads(result.output.strip())
        break
    except json.JSONDecodeError:
        prompt += "\n\nReturn ONLY valid JSON. No markdown, no prose."
else:
    raise ValueError("model failed to return JSON")

The for/else construct runs the else block when the loop completes without break — clean way to express "exhausted retries".

Common ways the model slips

Slip	Output
Markdown fence	```json\n{...}\n```
Prose preamble	`"Here's the JSON: {...}"`
Trailing commentary	`{...} - this classifies as positive.`
Smart quotes instead of straight	`{"sentiment": "positive"}` (different quote chars — still a parse error)

Mitigations beyond retry

Strict prompt — "Return ONLY valid JSON, no prose, no markdown" — works ~90% of the time.
Strip markdown fences before parsing — defensive: result.output.strip().lstrip('```json').rstrip('```').strip() — helps when the model uses fences but is otherwise valid.
pydantic_ai's typed output — output_type=YourModel handles parsing internally and retries automatically. The cleaner production answer.

When to give up

If 2-3 retries fail, the prompt itself is the problem. Either the task is too ambiguous, or the model is too small. Switch model or simplify the schema before adding more retries.

Parsing failures — `try/except json.JSONDecodeError`

Common ways the model slips

Mitigations beyond retry

When to give up

Parsing failures — `try/except json.JSONDecodeError`

Common ways the model slips

Mitigations beyond retry

When to give up

Parsing failures — `try/except json.JSONDecodeError`

Common ways the model slips

Mitigations beyond retry

When to give up

Sign up to practice

Parsing failures — `try/except json.JSONDecodeError`

Common ways the model slips

Mitigations beyond retry

When to give up

Sign up to practice

Parsing failures — try/except json.JSONDecodeError

Common ways the model slips

Mitigations beyond retry

When to give up

Parsing failures — try/except json.JSONDecodeError

Common ways the model slips

Mitigations beyond retry

When to give up

Parsing failures — try/except json.JSONDecodeError

Common ways the model slips

Mitigations beyond retry

When to give up

Sign up to practice

Parsing failures — try/except json.JSONDecodeError

Common ways the model slips

Mitigations beyond retry

When to give up

Sign up to practice

Parsing failures — `try/except json.JSONDecodeError`

Parsing failures — `try/except json.JSONDecodeError`

Parsing failures — `try/except json.JSONDecodeError`

Parsing failures — `try/except json.JSONDecodeError`