Three paths. Pick by your baseline, not your hope.
The 4-week plan is the canonical path. The 2-week sprint and 8-week comprehensive are calibrated alternates for candidates with very different starting baselines. Take the diagnostic first, then pick the path that matches your score and weekly hours, not the one that sounds aspirational.
Pick this if…
Each card lists the profile that fits, the weekly hour commitment, and the total time investment.
2-week sprint
Compressed pass through the 4-week plan: skim weeks 1-2, sprint weeks 3-4. Skip K-pages you already know. Concentrate on practice volume.
- You're already shipping production agents on Claude API or Claude Code
- You scored ≥75% on the diagnostic without preparation
- You can dedicate 4+ hours/day for 14 days
4-week canonical (Recommended)
The canonical recommended plan. Day-by-day across all four weeks: foundations + diagnostic, then D1+D2, then D3+D4, then D5 + scenarios + final mock.
- You ship code in production but haven't built agentic systems before
- You scored 50-75% on the diagnostic
- You can dedicate 1-2 hours/day on weekdays + a longer weekend session
8-week comprehensive
Spread the 4-week themes across 8 weeks: extra K-page reading time, more practice cycles, every concept page deeply read with the code samples followed.
- You're newer to Claude or LLM development overall
- You scored <50% on the diagnostic
- You prefer a slower pace with deeper Knowledge-pillar reading
Three rules.
- Diagnostic score is the tiebreaker. Take the free 10-Q diagnostic. ≥75% → 2-week or 4-week. 50-75% → 4-week. <50% → 8-week.
- Hours available wins ties. If you can't commit 10 hrs/week, do not pick the 4-week plan. The 8-week plan covers the same material at a sustainable pace.
- Production experience is the third lever. If you have shipped agentic systems on Claude API or Claude Code, you can compress. If you haven't, you cannot, regardless of test score.
The 2-week and 8-week paths are written as alternates inside the 4-week plan page, not as separate full pages. Open the 4-week plan and scroll to the alternates section for compression / expansion guidance.
How Long to Prepare: Daily Claude Code User vs. New to AI Engineering
Two starting baselines, two timelines, one shared midpoint milestone. Use the diagnostic score as your compression signal.
- How long should a daily Claude Code user prepare for the CCA-F?
- Two to three weeks is the typical window for candidates who are already fluent with tool use, context windows, and agentic loops day to day. Most of D1, D2, and D5 maps onto lived experience, so compression time should be spent on D3 (Agent Operations) and D4 (Prompt Engineering) edge cases where exam framing diverges from intuition.
- How long should someone new to AI engineering prepare for the CCA-F?
- Plan for six to eight weeks. Build foundational fluency in agentic architectures (D1, 27% of the exam) before drilling scenarios, and allocate a full week to tool design patterns (D2) - that domain demands conceptual load with no shortcut from prior hands-on work.
- What signal indicates a candidate can compress their study timeline?
- Score 80% or higher on the /practice/diagnostic without reviewing material first. Clearing that bar consistently is a reliable signal you can compress your planned timeline by 30 to 40%, regardless of starting background.
- What is the shared halfway-point milestone across both tracks?
- Both daily-user and newcomer tracks should complete at least one full /practice/mock under timed conditions at the midpoint, then review every wrong answer against its /concepts/[slug] page before continuing.
- What is the biggest over-preparation trap for experienced Claude Code users?
- Re-studying topics they already execute correctly in production. Use /exam-guide/anti-patterns to identify exam-specific failure modes that don't surface in day-to-day Claude Code usage - that's where the marginal points come from, not from rereading material you already apply.
Last updated