Learning Tracks

A track is a sequenced path through the mirrored courses and their starter projects — the syllabus that turns the catalog of individual lessons into an order you can actually follow. There are eight tracks, and they're meant to be done in order, Track 1 → Track 8, each one building on the last.

The on-ramps: if you're brand-new, start with a prerequisite track first. track-0a-non-technical is the gentle, no-code path for non-technical learners; track-0b-never-used-claude is the fast orientation for people comfortable with tech but new to Claude. Both feed into Track 1.

The whole path at a glance

flowchart TD
    A0A["Track 0A<br/>Non-technical on-ramp"] --> T1
    A0B["Track 0B<br/>Never used Claude"] --> T1
    T1["Track 1<br/>Claude Foundations"] --> T2["Track 2<br/>Effective Prompting"]
    T2 --> T3["Track 3<br/>Building with the API"]
    T3 --> T4["Track 4<br/>Claude Code"]
    T4 --> T5["Track 5<br/>MCP & Tools"]
    T5 --> T6["Track 6<br/>Agents, Skills & Subagents"]
    T6 --> T7["Track 7<br/>Deployment & Scale"]
    T7 --> T8["Track 8<br/>Certification (CCA-F)"]
    A0A -. "stop here is OK" .-> DONE["Everyday Claude user"]

Non-technical learners can stop after the on-ramp or after Track 1 and still have a complete, useful skill set. Developers aiming at the Claude Certified Architect — Foundations (CCA-F) exam run all eight tracks; Track 8 is the capstone that ties them to the exam domains.

All eight tracks

#	Track	Who it's for	Est. time	Courses	Projects
1	track-1-claude-foundations	Everyone, after the on-ramp	~4–5 h	claude-101, ai-fluency-framework-foundations, ai-capabilities-and-limitations	p01, p03
2	track-2-effective-prompting	Anyone who wants reliable outputs	~6–8 h	prompt-engineering-interactive-tutorial, real-world-prompting, prompt-evaluations	p02, p09
3	track-3-building-with-the-api	Developers	~10–14 h	building-with-the-claude-api	p05, p06, p07, p08
4	track-4-claude-code	Developers	~8–10 h	claude-code-101, claude-code-in-action, introduction-to-claude-cowork	p10, p11, p04
5	track-5-mcp-and-tools	Developers	~8–10 h	introduction-to-mcp, mcp-advanced-topics	p12, p15
6	track-6-agents-skills-subagents	Developers	~8–10 h	introduction-to-agent-skills, introduction-to-subagents	p13, p14, p15
7	track-7-deployment-and-scale	Developers / platform & ops	~6–8 h	claude-with-amazon-bedrock, claude-with-vertex-ai	review of p08 (batch/caching)
8	track-8-certification-cca-f	Exam candidates	~30–80 h	capstone + all 301 notes	p16

The AI-fluency electives — ai-fluency-for-students, ai-fluency-for-educators, teaching-ai-fluency, ai-fluency-for-nonprofits — aren't on the core spine. They're audience-specific variants of Track 1's fluency material; do whichever matches your role any time after Track 1.

Recommended order

On-ramp → track-0a-non-technical or track-0b-never-used-claude.
Track 1 → Track 2 → everyone. After Track 1 you can use Claude well; after Track 2 your prompts are reliable.
Track 3 → Track 4 → the developer core: the API, then Claude Code.
Track 5 → Track 6 → extend Claude with tools/MCP, then build agents and multi-agent systems.
Track 7 → put it on a cloud with governance.
Track 8 → only when Tracks 1–7 are done; it's the certification capstone.

Tracks 5 and 6 lightly overlap (both touch p15-multi-agent-orchestrator) and may be taken in either order; everything else is strictly linear.

How to use a track

Each track file gives you: a frontmatter header, a short intro, an ordered checklist linking its courses and projects with time budgets, "you're ready to move on when…" criteria, and a "See also." Work top to bottom, build every project (reading without building doesn't transfer — see how-to-use-hyperframes), and tick each item off in progress.