CCAR-P Readiness Assessment¶
A self-assessment against every CCAR-P objective in
registry.yaml, with readiness criteria and remediation paths. Take it at the start of your preparation and again in the final week.Important: nothing in this repository produces an official score. Practice-question percentages, mock-exam results and the self-ratings below are unofficial indicators only and do not equal the scaled score (100–1,000, pass at 720) reported by the real exam. Scaled scores are produced by the official exam's own psychometric process and cannot be derived from a raw percentage on our materials.
How to rate yourself¶
For each objective, give yourself a score from 0 to 3:
| Score | Meaning |
|---|---|
| 0 | I could not explain this to a colleague |
| 1 | I recognise the terms but could not make a design decision |
| 2 | I can choose between options and justify the choice for a typical scenario |
| 3 | I have done this in practice or in the labs, and can handle edge cases and explain trade-offs to stakeholders |
Be honest: a 2 means you could defend the decision in a design review.
Self-assessment by objective¶
Domain 1 — Solution Design & Architecture (17%)¶
| ID | Can I... | Score |
|---|---|---|
| CCAR-P-1.1 | Turn a vague business request into a value hypothesis, decide whether an LLM is the right tool, and apply the four agent criteria? | |
| CCAR-P-1.2 | Draw an end-to-end architecture with input controls, processing, output gates and a feedback loop? | |
| CCAR-P-1.3 | Choose between a single call, each workflow pattern and an agent, and justify it? | |
| CCAR-P-1.4 | Design an orchestrator–workers system with scoped tools, budgets, condensed hand-offs and tracing? | |
| CCAR-P-1.5 | Decompose a complex problem by function, data, time, risk and verification? | |
| CCAR-P-1.6 | Build a cost-per-completed-task business case with SLAs and sensitivity ranges? |
Domain 2 — Claude Models, Prompting & Context Engineering (13%)¶
| ID | Can I... | Score |
|---|---|---|
| CCAR-P-2.1 | Pick model tier and effort per route from eval evidence, and state current model IDs and request rules? | |
| CCAR-P-2.2 | Write a structured system prompt and a layered guardrail design, without relying on prefill or sampling parameters? | |
| CCAR-P-2.3 | Decide when to use zero-shot, few-shot or reasoning, and how adaptive thinking and effort change that? | |
| CCAR-P-2.4 | Produce a context budget and choose among compaction, context editing, sub-agents and retrieval? | |
| CCAR-P-2.5 | Lay out a cache-friendly prompt, list silent invalidators, and decide when to use Skills or modular prompts? |
Domain 3 — Integration (19%)¶
| ID | Can I... | Score |
|---|---|---|
| CCAR-P-3.1 | Audit an agent's tools for bloat and redesign them as narrow, strict, read/write-separated tools? | |
| CCAR-P-3.2 | Find identity and authorisation gaps (shared credentials, token passthrough, missing ACL filters)? | |
| CCAR-P-3.3 | Justify a configuration with eval accuracy, latency percentiles and cost against an SLO? | |
| CCAR-P-3.4 | Specify tracing, metrics, redaction and sampling for an agent at high volume? | |
| CCAR-P-3.5 | Choose chunking and indexing for different document types with metadata and ACLs? | |
| CCAR-P-3.6 | Match retrieval strategy (dense, hybrid, query tools, agentic, long context, live API) to data and queries? | |
| CCAR-P-3.7 | Choose between direct tools, CLI, MCP and agent-to-agent for a given consumer and trust boundary? | |
| CCAR-P-3.8 | Decide between monolithic context and progressive discovery, and configure tool search correctly? |
Domain 4 — Evaluation, Testing & Optimization (16%)¶
| ID | Can I... | Score |
|---|---|---|
| CCAR-P-4.1 | Define primary, secondary and guardrail metrics across accuracy, latency, cost, safety and security? | |
| CCAR-P-4.2 | Design a stratified, versioned dataset with splits and a calibrated mixed-method grading approach? | |
| CCAR-P-4.3 | Plan shadow, canary and A/B rollouts with pre-registered metrics and rollback triggers? | |
| CCAR-P-4.4 | Distinguish prompt failure, retrieval failure, hallucination type and model mismatch with specific diagnostics? | |
| CCAR-P-4.5 | Apply cost and latency levers in the right order and prove the result? | |
| CCAR-P-4.6 | Specify what to log per call and which dashboards and alerts to build? |
Domain 5 — Governance, Safety & Risk Management (14%)¶
| ID | Can I... | Score |
|---|---|---|
| CCAR-P-5.1 | Design layered guardrails where enforcement does not depend on the model complying? | |
| CCAR-P-5.2 | Threat-model an LLM system and list its failure modes with controls? | |
| CCAR-P-5.3 | Choose the oversight level by impact, reversibility and confidence, and design effective review? | |
| CCAR-P-5.4 | Map GDPR, HIPAA and FedRAMP obligations to architecture controls and evidence? | |
| CCAR-P-5.5 | Plan bias testing (slices, counterfactuals) and transparency measures? |
Domain 6 — Stakeholder Communication & Lifecycle Management (14%)¶
| ID | Can I... | Score |
|---|---|---|
| CCAR-P-6.1 | Run discovery across value, users, data, risk, quality and non-functionals? | |
| CCAR-P-6.2 | Present options with consequences to executives, risk teams and engineers, and write an ADR? | |
| CCAR-P-6.3 | Agree SLOs for availability, latency, quality and cost, and run feedback loops? | |
| CCAR-P-6.4 | Assemble a complete handover pack and implementation guide? | |
| CCAR-P-6.5 | Lead a solution through handoff, monitoring, iteration, model migration and retirement? |
Domain 7 — Developer Productivity & Operational Enablement (7%)¶
| ID | Can I... | Score |
|---|---|---|
| CCAR-P-7.1 | Configure Claude Code for a team with correct settings precedence, permissions, hooks, CLAUDE.md and MCP? | |
| CCAR-P-7.2 | Choose among CLAUDE.md, Skills, subagents, hooks, plugins, headless mode and the Agent SDK for a workflow? | |
| CCAR-P-7.3 | Use Claude for incident investigation and debug common Claude API and Claude Code problems? |
Readiness criteria¶
You are likely ready to book (or keep) your exam date when all of the following hold:
- Self-assessment: no objective scored 0; no more than three objectives scored 1; every objective in Domains 1, 3 and 4 (52% of the exam) scored 2 or higher.
- Practice questions: consistently at or above 80% on the CCAR-P practice question banks across all seven domains, on questions you have not seen before, and you can explain why each wrong option is wrong.
- Mock exams: at least two timed mock examinations at 63 items in 120 minutes, each finished with time to review and scoring 80% or higher.
- Deliverables: at least five of the ten practical exercises completed and self-reviewed at 8/10 or higher, including the ADR, eval plan and threat model.
- Scenarios: you reach the recommended decision on at least 13 of the 16 decision scenarios before reading the analysis.
These thresholds are our suggestion for building a safety margin. They are not a conversion to the official scaled score, and meeting them does not guarantee a pass.
Remediation paths¶
Find your weakest objectives and follow the matching path. Each path ends by re-scoring the objective.
Use the CCAR-P flashcards daily for recall, filtered by the objectives you scored 0 or 1. The coverage matrix shows which practice questions map to each objective.