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.

Memory Injection: Context via system_prompt — Ai For Makers

Day 12 · ~12 min●

You have a numbered fact list and a goal the user wants the agent to pursue. Where inside the call do you put the memory so the agent reads it first?

In the user prompt before the goal? Or is there a special slot for background knowledge?

There is — system_prompt. Anything passed as system_prompt arrives before the user message, framing how the model reads the goal. Build the prompt from the context string at call time:

python

agent = Agent(model, system_prompt=f"Known facts:\n{context}")
result = agent.run_sync(goal)
print(result.output)

So the system prompt is the agent's briefing, and each goal is a task that runs against that same briefing?

Exactly. The system_prompt sets the world; the user message sets the task. If the facts change, rebuild the agent. If the goal changes, keep the agent and call again. The full injector:

python

def agent_with_context(context: str, goal: str) -> str:
    agent = Agent(model, system_prompt=f"Known facts:{chr(10)}{context}")
    result = agent.run_sync(goal)
    return result.output

Why chr(10) again? Cannot I just use a plain newline inside the f-string?

Both work. Using chr(10) keeps the pattern consistent with the context builder from Day 10, so memory flows through the same character on both sides. In production you would pick one and stick with it.

And the same context can drive any goal I pass in — one briefing, many tasks?

One briefing, many tasks. This is how memory-augmented agents scale: the context carries what the agent knows, the goal carries what the user wants, and the model blends them into a grounded answer.

Context Injection via system_prompt

TL;DR: pass the fact list as system_prompt; pass the user's goal as the run_sync argument.

system_prompt — the agent's static briefing
run_sync(goal) — the per-call dynamic task
f-string build — f"Known facts:{chr(10)}{context}"
Rebuild on change — new facts mean a new agent instance

Where memory lives

Slot	Purpose	Lifetime
`system_prompt`	stable context	per agent
`run_sync(...)` argument	per-call goal	per call

Separating static and dynamic inputs keeps the agent's reasoning clean and reproducible.

Day 12 · ~12 min●

You have a numbered fact list and a goal the user wants the agent to pursue. Where inside the call do you put the memory so the agent reads it first?

In the user prompt before the goal? Or is there a special slot for background knowledge?

There is — system_prompt. Anything passed as system_prompt arrives before the user message, framing how the model reads the goal. Build the prompt from the context string at call time:

python

agent = Agent(model, system_prompt=f"Known facts:\n{context}")
result = agent.run_sync(goal)
print(result.output)

So the system prompt is the agent's briefing, and each goal is a task that runs against that same briefing?

Exactly. The system_prompt sets the world; the user message sets the task. If the facts change, rebuild the agent. If the goal changes, keep the agent and call again. The full injector:

python

def agent_with_context(context: str, goal: str) -> str:
    agent = Agent(model, system_prompt=f"Known facts:{chr(10)}{context}")
    result = agent.run_sync(goal)
    return result.output

Why chr(10) again? Cannot I just use a plain newline inside the f-string?

And the same context can drive any goal I pass in — one briefing, many tasks?

One briefing, many tasks. This is how memory-augmented agents scale: the context carries what the agent knows, the goal carries what the user wants, and the model blends them into a grounded answer.

Context Injection via system_prompt

TL;DR: pass the fact list as system_prompt; pass the user's goal as the run_sync argument.

system_prompt — the agent's static briefing
run_sync(goal) — the per-call dynamic task
f-string build — f"Known facts:{chr(10)}{context}"
Rebuild on change — new facts mean a new agent instance

Where memory lives

Slot	Purpose	Lifetime
`system_prompt`	stable context	per agent
`run_sync(...)` argument	per-call goal	per call

Separating static and dynamic inputs keeps the agent's reasoning clean and reproducible.

Context Injection via system_prompt

TL;DR: pass the fact list as system_prompt; pass the user's goal as the run_sync argument.

system_prompt — the agent's static briefing

run_sync(goal) — the per-call dynamic task

f-string build — f"Known facts:{chr(10)}{context}"

Rebuild on change — new facts mean a new agent instance

Where memory lives

Slot	Purpose	Lifetime
`system_prompt`	stable context	per agent
`run_sync(...)` argument	per-call goal	per call

Separating static and dynamic inputs keeps the agent's reasoning clean and reproducible.

Context Injection via system_prompt

TL;DR: pass the fact list as system_prompt; pass the user's goal as the run_sync argument.

system_prompt — the agent's static briefing

run_sync(goal) — the per-call dynamic task

f-string build — f"Known facts:{chr(10)}{context}"

Rebuild on change — new facts mean a new agent instance

Where memory lives

Slot	Purpose	Lifetime
`system_prompt`	stable context	per agent
`run_sync(...)` argument	per-call goal	per call

Separating static and dynamic inputs keeps the agent's reasoning clean and reproducible.

Context Injection via system_prompt

Where memory lives

Context Injection via system_prompt

Where memory lives

Context Injection via system_prompt

Where memory lives

Sign up to practice

Context Injection via system_prompt

Where memory lives

Sign up to practice