# Claude Certified

> Free, independent, community-built prep site for the Claude Certified Architect – Foundations (CCA-F) exam by Anthropic: 1,374 practice questions, 165 study-guide articles across the 5 exam domains, a full-length timed mock exam, and a blog answering the common questions about the exam. English and Simplified Chinese. Not affiliated with, sponsored by, or endorsed by Anthropic; Claude is a trademark of Anthropic, PBC.

Official exam facts as of mid-2026 (confirm on Anthropic / Pearson VUE before registering): 60 scenario-based questions, 120 minutes, US$125, passing scaled score 720/1000, credential valid 12 months, delivered by Pearson VUE. Registration and official training: https://anthropic.skilljar.com/

Full text of the blog and study guide for LLM consumption: https://certified-architect.com/llms-full.txt

## Core pages

- [Home](https://certified-architect.com/): Claude Certified Architect exam prep overview, domain weights, FAQ (cost, format, official vs this site, other Claude certifications).
- [CCA-F Study Guide](https://certified-architect.com/cca-f-study-guide): 165 study articles organised by exam domain and knowledge point.
- [Practice Questions](https://certified-architect.com/cca-f-practice-questions): 1,374 free practice questions, one per knowledge point, with explanations.
- [Claude Certified Architect Mock Exam](https://certified-architect.com/claude-certified-architect-mock-exam): 60-question, 90-minute timed simulation sampled by official domain weight, 72% pass mark.
- [CCA-F Cheat Sheet](https://certified-architect.com/cca-f-cheat-sheet): quick reference for API syntax, XML tags, prompt structures and decision trees.
- [Blog](https://certified-architect.com/blog): 40 articles on exam cost, format, passing score, study plans, domain deep dives, career value and technical how-tos.
- [About](https://certified-architect.com/about): what the site is, exam domain weights, disclaimer.
- [Community](https://certified-architect.com/community): shared study strategies and exam tips.

## Exam domains (study guide + practice per domain)

- [Domain 1 — Agentic Architecture & Orchestration (27%)](https://certified-architect.com/cca-f-study-guide/domain1): agentic loop, hub-and-spoke orchestration, sub-agent isolation, hooks vs prompt enforcement, decomposition. Practice: https://certified-architect.com/cca-f-practice-questions/domain/1
- [Domain 2 — Tool Design & MCP Integration (18%)](https://certified-architect.com/cca-f-study-guide/domain2): tool descriptions, isError, tool_choice, tool count, MCP scope and primitives. Practice: https://certified-architect.com/cca-f-practice-questions/domain/2
- [Domain 3 — Claude Code Configuration & Workflows (20%)](https://certified-architect.com/cca-f-study-guide/domain3): CLAUDE.md hierarchy, rules, skills, plan mode, headless -p, hooks. Practice: https://certified-architect.com/cca-f-practice-questions/domain/3
- [Domain 4 — Prompt Engineering & Structured Output (20%)](https://certified-architect.com/cca-f-study-guide/domain4): specific criteria, few-shot, tool_use schemas, retries, Batch API. Practice: https://certified-architect.com/cca-f-practice-questions/domain/4
- [Domain 5 — Context Management & Reliability (15%)](https://certified-architect.com/cca-f-study-guide/domain5): summarization loss, escalation triggers, structured errors, degradation, monitoring. Practice: https://certified-architect.com/cca-f-practice-questions/domain/5
- [Foundations primer](https://certified-architect.com/cca-f-study-guide/foundations): Messages API, tool use, MCP, Claude Code, Agent SDK, JSON Schema, prompting basics.

## Blog

### Exam facts

- [CCA Exam Passing Score: What 720 Out of 1000 Really Means](https://certified-architect.com/blog/cca-exam-passing-score): The CCA exam passing score is a scaled 720 on a 100 to 1000 scale. Learn why scaled scores are not percentages and how to read your mock results against it.
- [CCA Exam Retake Policy: Cost, Wait Time, and a Retake Plan](https://certified-architect.com/blog/cca-exam-retake-policy-cost-wait-time-strategy): The CCA exam retake policy is not fully published by Anthropic. Here is what to verify about cost and wait time, and a concrete study plan for a second attempt.
- [CCA Foundations Exam Guide 2026: Format, Fee, Domains, Plan](https://certified-architect.com/blog/cca-foundations-exam-guide-2026): A one-page CCA Foundations exam guide for 2026: what the CCA-F is, who it is for, format, fee, the five domains, and how to go from register to badge.
- [Claude Certified Architect Exam Cost: The $125 Fee Explained](https://certified-architect.com/blog/cca-exam-cost): The Claude Certified Architect exam cost is US$125 for the Foundations (CCA-F) exam. Here is what the fee covers, what it does not, and how to prep for free.
- [How Hard Is the CCA Exam? Judgment, Not Memory](https://certified-architect.com/blog/how-hard-is-cca-foundations-exam): How hard is the CCA exam? The CCA-F is hard on design judgment rather than recall. See which domains trip candidates up, why, and what makes the exam easier.
- [How Long Is the CCA Exam? 120 Minutes and a Pacing Plan](https://certified-architect.com/blog/how-long-is-cca-foundations-exam): How long is the CCA exam? The CCA-F gives you 120 minutes for 60 questions, about 2 minutes each. Here is a flag-and-return pacing plan that leaves a buffer.
- [How Many Questions on the CCA Exam? 60, Split by Domain](https://certified-architect.com/blog/how-many-questions-cca-exam): How many questions is the CCA exam? The CCA-F has 60 scenario-based questions in 120 minutes. See the per-domain split at official weights and a pacing plan.
- [Is the CCA Exam Multiple Choice? Format, Types, and a Sample](https://certified-architect.com/blog/is-cca-exam-multiple-choice): Is the CCA exam multiple choice? Yes: 60 scenario-based multiple-choice and multiple-response questions in 120 minutes. See the format and a sample question.

### Study strategy

- [10 CCA Exam Anti-Patterns the CCA-F Repeatedly Punishes](https://certified-architect.com/blog/cca-exam-anti-patterns): The CCA exam anti-patterns the CCA-F punishes again and again: prompt-only enforcement, uniform errors, every tool for every agent, unbounded loops and more.
- [CCA Domain Weighting: How to Split Study Time Across 5 Domains](https://certified-architect.com/blog/cca-domain-weighting-strategy-focus-study-time): Turn CCA domain weighting (27/18/20/20/15) into study hours for a 40-hour or 80-hour budget, why Domain 1 gets the most, and why Domain 5 is not skippable.
- [CCA Exam Day: What to Expect and How to Pace 60 Questions](https://certified-architect.com/blog/cca-exam-day-strategy-what-to-expect): CCA exam day guide for the CCA-F: Pearson VUE OnVUE check-in basics, pacing at 2 minutes a question, flag-and-return, multiple-response and least-bad answers.
- [CCA Exam Domains Explained: All 5 CCA-F Domains and Weights](https://certified-architect.com/blog/cca-foundations-exam-domains-explained): The 5 CCA exam domains explained: what Agentic Architecture, Tool Design & MCP, Claude Code, Prompt Engineering and Context Management really test.
- [CCA Study Plan: 30 Days to the CCA-F Exam, Week by Week](https://certified-architect.com/blog/cca-exam-study-schedule-30-day-plan): A 30-day CCA study plan for the CCA-F exam: week-by-week schedule across all 5 domains, weekly mock exams, linked study pages, plus a compressed 14-day variant.
- [Free CCA Practice Questions: 1,374 CCA-F Questions by Domain](https://certified-architect.com/blog/cca-foundations-exam-practice-questions-free): Free CCA practice questions for the CCA-F: what a good scenario question looks like, 1,374 questions by domain, per-knowledge-point mode vs timed mock, samples.
- [How to Become a Claude Certified Architect: Step-by-Step Path](https://certified-architect.com/blog/how-to-become-claude-certified-architect): How to become a Claude Certified Architect: prerequisites, the five CCA-F domains, practice, registering via Anthropic Academy and Pearson VUE, badge, renewal.

### Domain deep dives

- [CCA Domain 1: Agentic Architecture & Orchestration Explained](https://certified-architect.com/blog/agentic-architecture-orchestration-cca-domain-1): What CCA domain 1 agentic architecture covers, why it carries 27% of the CCA-F, the concepts tested, a worked scenario, and the traps that cost points.
- [CCA Domain 2: Tool Design & MCP Integration Explained](https://certified-architect.com/blog/tool-design-mcp-cca-domain-2): CCA domain 2 tool design and MCP explained: what the 18% domain covers, how descriptions, isError, tool count and MCP scope are tested.
- [CCA Domain 3: Claude Code Configuration & Workflows Explained](https://certified-architect.com/blog/claude-code-config-workflows-cca-domain-3): What CCA domain 3 Claude Code configuration tests: the CLAUDE.md hierarchy, rules and skills, plan mode and CI usage, with a worked team scenario and traps.
- [CCA Domain 4: Prompt Engineering & Structured Output Explained](https://certified-architect.com/blog/prompt-engineering-structured-output-cca-domain-4): What CCA domain 4 prompt engineering and structured output tests: specific criteria, few-shot, tool_use schemas, retries and Batch API.
- [CCA Domain 5: Context Management & Reliability Explained](https://certified-architect.com/blog/context-management-reliability-cca-domain-5): CCA domain 5 context management and reliability explained: summarization loss, escalation triggers, structured errors, degradation and monitoring.
- [CCA Study Tips for Domain 1: Agentic Architecture & Orchestration](https://certified-architect.com/blog/cca-study-tips-domain-1-agentic-architecture): CCA study tips for agentic architecture: a reading order through the Domain 1 articles, what to memorise vs reason about, cadence and self-checks.
- [CCA Study Tips for Domain 2: Tool Design & MCP Mastery](https://certified-architect.com/blog/cca-study-tips-domain-2-tool-design-mcp-mastery): CCA study tips tool design MCP: the reading order through the Domain 2 articles, what to memorise vs reason about, a practice cadence and self-checks.
- [CCA Study Tips for Domain 3: Claude Code Configuration & Workflows](https://certified-architect.com/blog/cca-study-tips-domain-3-claude-code-config-workflow): CCA study tips for Claude Code configuration: a reading order through the Domain 3 articles, what to memorise vs reason about, and drills.
- [CCA Study Tips for Domain 4: Prompt Engineering Mastery](https://certified-architect.com/blog/cca-study-tips-domain-4-prompt-engineering-mastery): CCA study tips for prompt engineering: a reading order through the Domain 4 articles, the prompt vs schema vs code split, and self-checks.
- [CCA Study Tips for Domain 5: Context Management & Reliability](https://certified-architect.com/blog/cca-study-tips-domain-5-context-management-reliability): CCA study tips for context management and reliability: Domain 5 reading order, what to memorise vs reason about, and self-checks.

### Career & value

- [AI Engineer Certifications 2026: The Landscape and Where CCA-F Fits](https://certified-architect.com/blog/ai-engineer-certifications-2026): AI engineer certifications 2026: a map of Anthropic's four Claude exams, cloud-vendor AI certs and vendor-neutral options, what each proves, and how to choose.
- [CCA vs AWS AI Practitioner: Which Certification Should You Take?](https://certified-architect.com/blog/cca-vs-aws-ai-practitioner-which-certification): CCA vs AWS AI Practitioner: Claude-specific agent architecture versus cloud AI literacy, what each exam covers, who should take which, and when both make sense.
- [CCA-F vs AI Bootcamp: $125 Certification or Thousand-Dollar Course?](https://certified-architect.com/blog/cca-f-vs-ai-bootcamp): CCA vs bootcamp: a $125 Claude certification with free preparation against multi-thousand-dollar AI bootcamps, what each proves, who each fits, how to combine.
- [Claude Architect Responsibilities: A Week in the Role, by Exam Domain](https://certified-architect.com/blog/cca-real-world-reports-responsibilities-salaries): Claude architect responsibilities in practice: agent loops, tool boundaries, MCP servers, Claude Code workflows and evals, mapped to CCA-F domains, plus pay.
- [Claude Certified Architect Jobs: Titles, Skills and Domain Mapping](https://certified-architect.com/blog/cca-architect-jobs): Claude Certified Architect jobs rarely carry that title. See which roles use CCA-F skills, what job descriptions ask for, and how the five domains map.
- [Claude Certified Architect Salary and Career Outlook (2026)](https://certified-architect.com/blog/claude-certified-architect-salary-career-2026): Claude Certified Architect salary: Anthropic publishes no pay data for the CCA-F. Here is what the credential signals, the roles it maps to, and how to use it.
- [How to Prepare for the CCA-F as a Career Switch](https://certified-architect.com/blog/how-to-prepare-cca-f-career-switch): A CCA career switch plan for engineers from non-AI backgrounds and non-developers: what to learn first, realistic timelines, and portfolio plus badge.
- [Is Claude Certified Architect Worth It? An Honest Cost-Benefit Check](https://certified-architect.com/blog/claude-certified-architect-certification-worth-it): Is Claude Certified Architect worth it? Weigh the $125 fee and 40–80 study hours against what the CCA-F actually proves, who benefits, and what to do instead.
- [Who Should Get the CCA Foundations Certification (and Who Should Skip)](https://certified-architect.com/blog/who-should-get-cca-foundations-certification): Who should get CCA Foundations? Five personas it fits, three it does not, and the practical prerequisites for the Claude Certified Architect – Foundations exam.
- [Why Claude Certification Matters (and Where It Stops Mattering)](https://certified-architect.com/blog/why-claude-certification-is-important): Why Claude certification is worth taking seriously: a verifiable signal, a shared team vocabulary, a forcing function for architecture, and its limits.

### Technical guides

- [9 MCP Server Design Mistakes and How to Avoid Them](https://certified-architect.com/blog/common-mcp-server-design-mistakes-how-to-avoid): MCP server design mistakes that break agents in production and on the CCA-F exam: vague descriptions, uniform errors, missing isError, tool sprawl, and fixes.
- [Claude Code Hooks vs CLAUDE.md Instructions: Which Enforces What](https://certified-architect.com/blog/claude-code-hooks-vs-claude-md-instructions): Claude Code hooks vs CLAUDE.md instructions: when a rule must be a PreToolUse or PostToolUse hook, when a prompt is enough, and the decision table CCA-F uses.
- [Claude Sub-Agents and Context Isolation, Explained](https://certified-architect.com/blog/claude-sub-agents-context-isolation-explained): Claude sub-agents context isolation explained: why each sub-agent starts empty, hub-and-spoke handoffs, parallel spawning, and when to fork vs start fresh.
- [How to Learn Claude Agent Development: A 7-Step Path](https://certified-architect.com/blog/how-to-learn-claude-agent-development): A practical path to learn Claude agent development: Messages API, tool use, agentic loop, sub-agents, MCP, Claude Code, Agent SDK, with CCA-F as a milestone.
- [How to Write CLAUDE.md: A Practical Guide for Claude Code Teams](https://certified-architect.com/blog/how-to-write-effective-claude-md-file): How to write CLAUDE.md files Claude Code actually follows: the three-level hierarchy, what belongs in the file, imports, path-scoped rules, and anti-patterns.

## Study guide articles

### Foundations

- [F1.1 The Messages API Is Stateless — Every Request Stands Alone](https://certified-architect.com/cca-f-study-guide/foundations/f1-1-messages-api-structure): Messages API Request Structure
- [F1.2 Two Roles, No Exceptions](https://certified-architect.com/cca-f-study-guide/foundations/f1-2-message-roles): Message Roles
- [F1.3 Four Content Block Types, Each With a Direction](https://certified-architect.com/cca-f-study-guide/foundations/f1-3-content-block-types): Content Block Types
- [F1.4 Reading the Response: stop_reason Is Your Control Signal](https://certified-architect.com/cca-f-study-guide/foundations/f1-4-response-structure): Response Structure
- [F1.5 Tokens: The Currency of Every API Call](https://certified-architect.com/cca-f-study-guide/foundations/f1-5-tokens-and-context-window): Tokens and Context Window
- [F2.1 The Tool Use Dance: Six Steps, Three API Calls](https://certified-architect.com/cca-f-study-guide/foundations/f2-1-tool-use-complete-flow): Tool Use Complete Flow
- [F2.2 Defining Tools: Name, Description, and Input Schema](https://certified-architect.com/cca-f-study-guide/foundations/f2-2-tool-definition-format): Tool Definition Format
- [F2.3 The tool_use Block: What the Model Sends When It Wants a Tool Called](https://certified-architect.com/cca-f-study-guide/foundations/f2-3-tool-use-content-block): tool_use Content Block
- [F2.4 Sending Tool Results Back: The tool_result Block](https://certified-architect.com/cca-f-study-guide/foundations/f2-4-tool-result-content-block): tool_result Content Block
- [F2.5 Virtual Tools: Hijacking tool_use for Guaranteed JSON](https://certified-architect.com/cca-f-study-guide/foundations/f2-5-tool-use-as-structured-output): tool_use as Structured Output
- [F3.1 MCP: One Protocol to Connect Them All](https://certified-architect.com/cca-f-study-guide/foundations/f3-1-what-is-mcp): What is MCP
- [F3.2 Tools, Resources, and Prompts: MCP's Three Primitives](https://certified-architect.com/cca-f-study-guide/foundations/f3-2-mcp-three-primitives): MCP Three Primitives
- [F3.3 The camelCase Trap: MCP Tool Definitions vs Claude API](https://certified-architect.com/cca-f-study-guide/foundations/f3-3-mcp-tool-definition-format): MCP Tool Definition Format
- [F3.4 Don't Disguise Errors as Empty Results](https://certified-architect.com/cca-f-study-guide/foundations/f3-4-mcp-call-tool-result): MCP CallToolResult
- [F3.5 stdio for Local, HTTP for Remote: Choosing MCP Transports](https://certified-architect.com/cca-f-study-guide/foundations/f3-5-mcp-server-transport-methods): MCP Server Transport Methods
- [F4.1 Claude Code: A CLI Tool, Not an IDE](https://certified-architect.com/cca-f-study-guide/foundations/f4-1-what-is-claude-code): What is Claude Code
- [F4.2 Grep Searches Content, Glob Searches Paths](https://certified-architect.com/cca-f-study-guide/foundations/f4-2-builtin-tools): Built-in Tools
- [F4.3 Three Levels of CLAUDE.md: User, Project, Directory](https://certified-architect.com/cca-f-study-guide/foundations/f4-3-configuration-hierarchy): Configuration Hierarchy
- [F4.4 Four Permission Modes: From Full Control to Full Automation](https://certified-architect.com/cca-f-study-guide/foundations/f4-4-permission-modes): Permission Modes
- [F5.1 Agent SDK: Claude Code's Agentic Powers, Programmable](https://certified-architect.com/cca-f-study-guide/foundations/f5-1-what-is-agent-sdk): What is Agent SDK
- [F5.2 query() vs ClaudeSDKClient: Stateless vs Stateful](https://certified-architect.com/cca-f-study-guide/foundations/f5-2-core-interfaces): Core Interfaces
- [F5.3 ClaudeAgentOptions: Configuring What an Agent Can Do](https://certified-architect.com/cca-f-study-guide/foundations/f5-3-claude-agent-options): ClaudeAgentOptions
- [F5.4 When Statefulness Actually Matters](https://certified-architect.com/cca-f-study-guide/foundations/f5-4-query-vs-claude-sdk-client): query vs ClaudeSDKClient
- [F5.5 Sub-Agents: Coordinator Delegates, Specialists Execute](https://certified-architect.com/cca-f-study-guide/foundations/f5-5-sub-agents-in-agent-sdk): Sub-Agents in Agent SDK
- [F6.1 Seven Types: string, number, integer, boolean, array, object, null](https://certified-architect.com/cca-f-study-guide/foundations/f6-1-json-schema-core-types): JSON Schema Core Types
- [F6.2 required, nullable, and enum: Schema Features That Shape Output](https://certified-architect.com/cca-f-study-guide/foundations/f6-2-schema-features): Schema Features
- [F7.1 System Prompt: Persistent Frame, Not Per-Turn Input](https://certified-architect.com/cca-f-study-guide/foundations/f7-1-system-prompt-vs-user-message): System Prompt vs User Message
- [F7.2 Few-Shot: Show, Don't Just Tell](https://certified-architect.com/cca-f-study-guide/foundations/f7-2-few-shot-vs-zero-shot): Few-shot vs Zero-shot
- [F7.3 Temperature: Determinism Is Not Accuracy](https://certified-architect.com/cca-f-study-guide/foundations/f7-3-temperature): Temperature
- [F8.1 Agents Do Things; Chatbots Talk About Things](https://certified-architect.com/cca-f-study-guide/foundations/f8-1-agent-vs-chatbot): Agent vs Chatbot
- [F8.2 The Agentic Loop: Reason → Act → Observe → Repeat](https://certified-architect.com/cca-f-study-guide/foundations/f8-2-agentic-loop): Agentic Loop
- [F8.3 Five Orchestration Patterns: From Simple Chains to Dynamic Coordination](https://certified-architect.com/cca-f-study-guide/foundations/f8-3-orchestration-patterns): Orchestration Patterns

### Agentic Architecture & Orchestration

- [K1.1.1 stop_reason Is the Only Loop Control That Matters](https://certified-architect.com/cca-f-study-guide/domain1/k1-1-1-agentic-loop-lifecycle): Agentic loop lifecycle & stop_reason
- [K1.1.2 The API Is Stateless: Send Full History Every Time](https://certified-architect.com/cca-f-study-guide/domain1/k1-1-2-tool-result-append): Tool result append to conversation history
- [K1.1.3 Model-Driven Decisions vs Hardcoded Decision Trees](https://certified-architect.com/cca-f-study-guide/domain1/k1-1-3-model-driven-vs-decision-tree): Model-driven vs pre-configured decision tree
- [S1.1.1 Three Anti-Patterns That Break Agentic Loops](https://certified-architect.com/cca-f-study-guide/domain1/s1-1-1-agentic-loop-anti-patterns): Agentic loop anti-patterns
- [K1.2.1 Hub-and-Spoke: All Roads Go Through the Coordinator](https://certified-architect.com/cca-f-study-guide/domain1/k1-2-1-hub-and-spoke-orchestration): Hub-and-spoke orchestration pattern
- [K1.2.2 Sub-Agents See Nothing: Context Must Be Explicitly Passed](https://certified-architect.com/cca-f-study-guide/domain1/k1-2-2-sub-agent-context-isolation): Sub-agent context isolation
- [K1.2.3 Narrow Decomposition Means Missing Coverage](https://certified-architect.com/cca-f-study-guide/domain1/k1-2-3-task-decomposition-scope): Task decomposition scope risks
- [S1.2.1 Don't Run Every Agent for Every Query](https://certified-architect.com/cca-f-study-guide/domain1/s1-2-1-dynamic-sub-agent-selection): Dynamic sub-agent selection
- [S1.2.2 Evaluate, Identify Gaps, Fill Them: The Iterative Refinement Loop](https://certified-architect.com/cca-f-study-guide/domain1/s1-2-2-iterative-refinement-loop): Iterative refinement loop
- [K1.3.1 Task Tool, allowedTools, and the Full AgentDefinition Configuration Surface](https://certified-architect.com/cca-f-study-guide/domain1/k1-3-1-task-tool-allowed-tools): Task tool & allowedTools config
- [K1.3.2 AgentDefinition: Every Field Matters, Nothing Should Be Left to Defaults](https://certified-architect.com/cca-f-study-guide/domain1/k1-3-2-agent-definition-config): AgentDefinition configuration
- [K1.3.3 Task Tool: Three Inputs, Four Response Fields, One Common Bug](https://certified-architect.com/cca-f-study-guide/domain1/k1-3-3-task-tool-invocation): Task tool complete definition
- [K1.3.4 Resume Saves 85% Tokens and Chains Across Multiple Follow-Ups](https://certified-architect.com/cca-f-study-guide/domain1/k1-3-4-sub-agent-resume): Sub-agent resume pattern
- [K1.3.5 Fork Eliminates Anchoring Bias: 89% vs 62% Accuracy](https://certified-architect.com/cca-f-study-guide/domain1/k1-3-5-fork-based-sessions): Fork-based session management
- [S1.3.1 Structured Data Preserves Attribution; Plain Text Destroys It](https://certified-architect.com/cca-f-study-guide/domain1/s1-3-1-context-passing-best-practices): Context passing best practices
- [S1.3.2 Parallel Execution: 80% Wall-Clock Reduction, Zero Extra Cost](https://certified-architect.com/cca-f-study-guide/domain1/s1-3-2-parallel-sub-agent-spawning): Parallel sub-agent spawning
- [S1.3.3 Goals + Standards Beat Step-by-Step: 82 vs 68 Quality Score](https://certified-architect.com/cca-f-study-guide/domain1/s1-3-3-coordinator-prompt-design): Coordinator prompt design
- [K1.4.1 The Model Will Override Your Instructions 4-15% of the Time](https://certified-architect.com/cca-f-study-guide/domain1/k1-4-1-programmatic-vs-prompt-enforcement): Programmatic enforcement vs prompt guidance
- [K1.4.2 Missing Customer Expectation Drops Satisfaction 23 Points](https://certified-architect.com/cca-f-study-guide/domain1/k1-4-2-structured-handoff-protocol): Structured handoff protocol
- [S1.4.1 4+ Concerns? Only 38% Get Fully Addressed Without Explicit Decomposition](https://certified-architect.com/cca-f-study-guide/domain1/s1-4-1-multi-concern-decomposition): Multi-concern request decomposition
- [K1.5.1 Five Hook Types Across the Full Session Lifecycle](https://certified-architect.com/cca-f-study-guide/domain1/k1-5-1-hook-types-and-triggers): Hook types and triggers
- [K1.5.2 HookMatcher: Targeted Matchers Save 40% Hook Processing Time](https://certified-architect.com/cca-f-study-guide/domain1/k1-5-2-hook-matcher-registration): Hook registration with HookMatcher
- [K1.5.3 PreToolUse Returns: deny, allow, modify — and the Bugs Between Them](https://certified-architect.com/cca-f-study-guide/domain1/k1-5-3-pre-tool-use-return-values): PreToolUse hook return values
- [S1.5.1 PostToolUse Normalization: 12% → 0.3% Date Errors, 90% Token Reduction](https://certified-architect.com/cca-f-study-guide/domain1/s1-5-1-post-tool-use-normalization): PostToolUse data normalization
- [S1.5.2 Select by Consequence AND Verifiability, Not by Compliance Rate](https://certified-architect.com/cca-f-study-guide/domain1/s1-5-2-hook-vs-prompt-selection): Hook vs prompt selection principle
- [K1.6.1 Dynamic Adds 30% Overhead on Fixed Tasks: Match Strategy to Task Type](https://certified-architect.com/cca-f-study-guide/domain1/k1-6-1-fixed-vs-dynamic-decomposition): Fixed pipeline vs dynamic decomposition
- [K1.6.2 Single-Pass Reviews Produce Contradictory Findings: Split Into Per-File + Cross-File](https://certified-architect.com/cca-f-study-guide/domain1/k1-6-2-multi-pass-code-review): Large code review multi-pass split
- [S1.6.1 Without Exploration First, 80% of Effort Goes to Trivial Code](https://certified-architect.com/cca-f-study-guide/domain1/s1-6-1-open-ended-task-decomposition): Open-ended task decomposition
- [K1.7.1 Resume Saves 44% Time and 58% Tokens in CI Pipelines](https://certified-architect.com/cca-f-study-guide/domain1/k1-7-1-named-session-resume): Named session resume
- [K1.7.3 The Agent Can't Detect Stale Data: You Must Decide Resume vs Fresh](https://certified-architect.com/cca-f-study-guide/domain1/k1-7-3-resume-vs-fresh-session): Resume vs fresh session + summary injection

### Tool Design & MCP Integration

- [K2.1.1 30% Misrouting with Minimal Descriptions, Near-Zero with Detailed Ones](https://certified-architect.com/cca-f-study-guide/domain2/k2-1-1-tool-description-selection): Tool description as primary selection mechanism
- [K2.1.2 Data-Driven Description Fixing: Target the High-Misrouting Pairs](https://certified-architect.com/cca-f-study-guide/domain2/k2-1-2-description-overlap-misrouting): Description overlap causes misrouting
- [K2.1.3 Removing Keyword Rules Improved Routing from 78% to 94%](https://certified-architect.com/cca-f-study-guide/domain2/k2-1-3-system-prompt-keyword-sensitivity): System prompt keyword sensitivity
- [S2.1.1 Three Fields, All Critical: name, description, input_schema](https://certified-architect.com/cca-f-study-guide/domain2/s2-1-1-tool-definition-format): Tool definition standard format
- [K2.2.1 Structured Errors: 78-95% Recovery vs 15% for Generic "Operation Failed"](https://certified-architect.com/cca-f-study-guide/domain2/k2-2-1-iserror-flag): MCP isError flag in CallToolResult
- [K2.2.2 Without Classification, the Agent Wastes 5 Retries on a Permission Error](https://certified-architect.com/cca-f-study-guide/domain2/k2-2-2-error-type-classification): Error type classification
- [K2.2.3 "Operation Failed" × 5 Retries × 30 Seconds = Customer Waiting for Nothing](https://certified-architect.com/cca-f-study-guide/domain2/k2-2-3-uniform-error-anti-pattern): Uniform error response anti-pattern
- [K2.2.4 "Your Order Doesn't Exist" — But It Does, the Database Was Just Down](https://certified-architect.com/cca-f-study-guide/domain2/k2-2-4-access-failure-vs-valid-empty): Access failure vs valid empty result
- [S2.2.1 Retry Locally First, Propagate with Context If Recovery Fails](https://certified-architect.com/cca-f-study-guide/domain2/s2-2-1-local-recovery-vs-propagation): Local recovery vs coordinator propagation
- [K2.3.1 3 Tools = 97% Accuracy, 18 Tools = 51%: The Tool Count Curve](https://certified-architect.com/cca-f-study-guide/domain2/k2-3-1-tool-count-selection-reliability): Tool count vs selection reliability
- [K2.3.2 tool_choice: auto, any, tool, none](https://certified-architect.com/cca-f-study-guide/domain2/k2-3-2-tool-choice-four-modes): tool_choice four modes
- [K2.3.3 Least Privilege: Each Agent Gets Only Its Role's Tools](https://certified-architect.com/cca-f-study-guide/domain2/k2-3-3-cross-role-scoped-tools): Cross-role scoped tools
- [K2.4.1 Project Scope in .mcp.json, User Scope in ~/.claude.json](https://certified-architect.com/cca-f-study-guide/domain2/k2-4-1-mcp-server-scope): MCP server project vs user scope
- [K2.4.2 ${ENV_VAR}: Keep Secrets Out of Version Control](https://certified-architect.com/cca-f-study-guide/domain2/k2-4-2-env-var-expansion): Environment variable expansion
- [K2.4.3 MCP Resources: Browse Content Without Tool Calls](https://certified-architect.com/cca-f-study-guide/domain2/k2-4-3-resources-vs-tools): MCP Resources vs Tools
- [K2.4.4 All MCP Servers Connect at Startup](https://certified-architect.com/cca-f-study-guide/domain2/k2-4-4-mcp-auto-discovery): MCP server auto-discovery
- [S2.4.1 Use Community Servers for Standard Integrations](https://certified-architect.com/cca-f-study-guide/domain2/s2-4-1-community-vs-custom-servers): Community vs custom MCP servers
- [K2.5.1 Each Built-in Tool Has One Job](https://certified-architect.com/cca-f-study-guide/domain2/k2-5-1-builtin-tool-responsibilities): Built-in tool responsibilities
- [K2.5.2 When Edit Fails: Read + Write as Fallback](https://certified-architect.com/cca-f-study-guide/domain2/k2-5-2-edit-failure-fallback): Edit failure fallback
- [S2.5.1 Grep Entry Points, Read Key Files, Trace Imports](https://certified-architect.com/cca-f-study-guide/domain2/s2-5-1-incremental-codebase-understanding): Incremental codebase understanding
- [S2.5.2 Grep Each Export to Find All Cross-Module References](https://certified-architect.com/cca-f-study-guide/domain2/s2-5-2-cross-module-function-tracing): Cross-module function tracing

### Claude Code Configuration

- [K3.1.1 Put Team Standards Where the Team Can Actually Get Them](https://certified-architect.com/cca-f-study-guide/domain3/k3-1-1-three-level-config-hierarchy): Three-level config hierarchy
- [K3.1.2 Mentioning a File Is Not the Same as Loading It](https://certified-architect.com/cca-f-study-guide/domain3/k3-1-2-import-syntax): @import syntax
- [K3.1.3 Twelve Unscoped Rules Files Waste 78% of Your Context Budget](https://certified-architect.com/cca-f-study-guide/domain3/k3-1-3-rules-directory): .claude/rules/ modular rules
- [S3.1.1 /memory Tells You What Is Actually Loaded — MEMORY.md Does Not](https://certified-architect.com/cca-f-study-guide/domain3/s3-1-1-memory-command-diagnosis): /memory command diagnosis
- [K3.2.1 Commands Are Prompts. Skills Are Configured Workflows.](https://certified-architect.com/cca-f-study-guide/domain3/k3-2-1-commands-vs-skills): Commands vs Skills
- [K3.2.2 Three Fields That Turn a Prompt Into a Safe Workflow](https://certified-architect.com/cca-f-study-guide/domain3/k3-2-2-skill-frontmatter): SKILL.md frontmatter
- [K3.2.3 Fork Isolates Memory, Not the Filesystem](https://certified-architect.com/cca-f-study-guide/domain3/k3-2-3-context-fork-isolation): context: fork isolation
- [K3.2.4 Personal Skills Live in ~/.claude/skills/ — Name Them Differently](https://certified-architect.com/cca-f-study-guide/domain3/k3-2-4-personal-skill-customization): Personal skill customization
- [S3.2.1 CLAUDE.md Is Always-On. Skills Are On-Demand. Do Not Mix Them Up.](https://certified-architect.com/cca-f-study-guide/domain3/s3-2-1-skills-vs-claude-md-selection): Skills vs CLAUDE.md selection
- [K3.3.1 86% of Your Rules Context Is Wasted Without Path Globs](https://certified-architect.com/cca-f-study-guide/domain3/k3-3-1-yaml-frontmatter-paths): YAML frontmatter paths glob
- [K3.3.2 Section Headers Do Not Enforce Scope — Path Rules Do](https://certified-architect.com/cca-f-study-guide/domain3/k3-3-2-path-rules-reduce-irrelevant-context): Path rules reduce irrelevant context
- [K3.3.3 File TYPE or File LOCATION — That Decides Which Mechanism to Use](https://certified-architect.com/cca-f-study-guide/domain3/k3-3-3-glob-rules-vs-directory-claude-md): Glob rules vs subdirectory CLAUDE.md
- [K3.4.1 55% Rollback Rate When You Skip Planning on Complex Tasks](https://certified-architect.com/cca-f-study-guide/domain3/k3-4-1-plan-mode-safe-exploration): Plan mode for safe exploration
- [K3.4.2 If You Already Know the Fix, Just Fix It](https://certified-architect.com/cca-f-study-guide/domain3/k3-4-2-direct-execution-scoped-changes): Direct execution for scoped changes
- [K3.4.3 120 Files Explored, 12% Context Used](https://certified-architect.com/cca-f-study-guide/domain3/k3-4-3-explore-subagent-context-isolation): Explore subagent for context isolation
- [S3.4.1 Plan → Explore → Direct: The Three-Phase Workflow for Complex Tasks](https://certified-architect.com/cca-f-study-guide/domain3/s3-4-1-plan-explore-direct-combo): Plan + Explore + Direct combo
- [K3.5.1 Text Descriptions Hit an Ambiguity Ceiling at 63%. Examples Reach 91%.](https://certified-architect.com/cca-f-study-guide/domain3/k3-5-1-concrete-examples-over-text): Concrete examples > text descriptions
- [K3.5.2 Tests First: 3 Iterations to 15/15. Implementation First: 7 Iterations to 12/15.](https://certified-architect.com/cca-f-study-guide/domain3/k3-5-2-test-driven-iteration): Test-driven iteration
- [K3.5.3 Three Rounds of Rework Because Nobody Asked About Deployment Topology](https://certified-architect.com/cca-f-study-guide/domain3/k3-5-3-interview-mode): Interview mode
- [K3.5.4 Classify Before Submitting: Interacting Together, Independent Separately](https://certified-architect.com/cca-f-study-guide/domain3/k3-5-4-interacting-vs-independent-issues): Interacting vs independent issues
- [K3.6.1 Without -p, Your CI Job Hangs Until Timeout Kills It](https://certified-architect.com/cca-f-study-guide/domain3/k3-6-1-p-flag-non-interactive-ci): -p flag for non-interactive CI
- [K3.6.2 Prompt Says "Output JSON" → 73%. Schema Flag → 99%.](https://certified-architect.com/cca-f-study-guide/domain3/k3-6-2-structured-ci-output): Structured CI output
- [K3.6.3 CLAUDE.md Is Your CI Agent's Only Source of Project Standards](https://certified-architect.com/cca-f-study-guide/domain3/k3-6-3-claude-md-ci-context): CLAUDE.md as CI project context
- [K3.6.4 Claude Reviewing Its Own Code Is Barely a Review](https://certified-architect.com/cca-f-study-guide/domain3/k3-6-4-session-isolation-self-review): Session isolation for unbiased review
- [S3.6.1 62% of Review Comments Were Duplicates. Prior Context Dropped It to 4%.](https://certified-architect.com/cca-f-study-guide/domain3/s3-6-1-incremental-review): Incremental review
- [S3.6.2 60% Duplicate Test Suggestions Because Nobody Showed Claude the Existing Tests](https://certified-architect.com/cca-f-study-guide/domain3/s3-6-2-test-generation-dedup): Test generation dedup

### Prompt Engineering & Optimization

- [K4.1.1 "Be Conservative" Means Nothing — 47% Agreement. "Lacks Sample Size" Means Something — 94%.](https://certified-architect.com/cca-f-study-guide/domain4/k4-1-1-specific-criteria-over-vague): Specific criteria > vague instructions
- [K4.1.2 One 60% False Positive Category Made Developers Ignore the 95% Accurate One](https://certified-architect.com/cca-f-study-guide/domain4/k4-1-2-false-positives-poison-trust): False positives poison trust
- [K4.1.3 Labels Alone: 41%. Text Conditions: 72%. Text + Code Examples: 94%.](https://certified-architect.com/cca-f-study-guide/domain4/k4-1-3-severity-definitions-with-examples): Severity definitions with examples
- [K4.2.1 Text Instructions Failed at 64%. Two Examples Reached 91%.](https://certified-architect.com/cca-f-study-guide/domain4/k4-2-1-few-shot-ultimate-calibration): Few-shot as ultimate calibration
- [K4.2.2 Three Diverse Examples Beat Eight Homogeneous Ones](https://certified-architect.com/cca-f-study-guide/domain4/k4-2-2-few-shot-design-principles): Few-shot design principles
- [K4.2.3 Every Example Shows All Fields Populated — So the Model Fabricates Missing Ones](https://certified-architect.com/cca-f-study-guide/domain4/k4-2-3-few-shot-reducing-hallucination): Few-shot reducing hallucination
- [S4.2.1 Paired REPORT + SKIP Examples: 67% → 94% Boundary Accuracy](https://certified-architect.com/cca-f-study-guide/domain4/s4-2-1-few-shot-format-consistency): Few-shot for format consistency
- [K4.3.1 tool_use Eliminates Structural Errors. Semantic Errors Remain.](https://certified-architect.com/cca-f-study-guide/domain4/k4-3-1-tool-use-guaranteed-structure): tool_use for guaranteed structure
- [K4.3.2 Forced, Any, Auto — Three Modes, Three Guarantees, Three Failure Patterns](https://certified-architect.com/cca-f-study-guide/domain4/k4-3-2-input-schema-enforcement): input_schema enforcement
- [K4.3.3 Required Fields on Optional Data: The #1 Structural Cause of Hallucination](https://certified-architect.com/cca-f-study-guide/domain4/k4-3-3-nullable-fields-optional-data): Nullable fields for optional data
- [K4.3.4 Without Format Rules, the Same Date Comes Out Three Different Ways](https://certified-architect.com/cca-f-study-guide/domain4/k4-3-4-closed-enum-format-standardization): Closed enum with "other" escape / Format standardization
- [K4.4.1 Blind Retry: 12% Fixed After 3 Attempts. Error Feedback: 87% Fixed After 1.](https://certified-architect.com/cca-f-study-guide/domain4/k4-4-1-semantic-validation-beyond-schema): Semantic validation beyond schema (retry with corrective feedback)
- [K4.4.2 Retrying Absent Data Causes Hallucination — Format Errors Converge, Missing Data Diverges](https://certified-architect.com/cca-f-study-guide/domain4/k4-4-2-retry-with-corrective-feedback): Retry with corrective feedback (error classification)
- [K4.4.3 Track Which Patterns Developers Dismiss — Then Add SKIP Examples](https://certified-architect.com/cca-f-study-guide/domain4/k4-4-3-confidence-scoring-routing): Confidence scoring for routing (feedback loop)
- [K4.4.4 Schema Says Valid. Line Items Don't Sum to Total. Both Are True.](https://certified-architect.com/cca-f-study-guide/domain4/k4-4-4-cross-field-consistency): Cross-field consistency validation
- [K4.5.1 50% Cost Savings — But Up to 24 Hours Wait](https://certified-architect.com/cca-f-study-guide/domain4/k4-5-1-batch-api-for-scale): Batch API for scale
- [K4.5.2 Results Return in Arbitrary Order — custom_id Is Your Only Correlation](https://certified-architect.com/cca-f-study-guide/domain4/k4-5-2-custom-id-result-correlation): custom_id for result correlation
- [K4.5.3 940 Succeeded, 45 Errored, 15 Expired — Resubmit Only the 60](https://certified-architect.com/cca-f-study-guide/domain4/k4-5-3-batch-error-handling): Batch error handling
- [K4.5.4 Poll, Download, Store Locally — Results Expire in 29 Days](https://certified-architect.com/cca-f-study-guide/domain4/k4-5-4-batch-size-timeout): Batch size and timeout
- [K4.5.5 Different Errors Need Different Fixes — Don't Resubmit the Whole Batch](https://certified-architect.com/cca-f-study-guide/domain4/k4-5-5-cost-optimization-batch): Cost optimization with batch
- [S4.5.1 Always Test on a Sample Before Full Batch — 18% Failure vs 3%](https://certified-architect.com/cca-f-study-guide/domain4/s4-5-1-batch-workflow-design): Batch workflow design
- [K4.6.1 Same-Session Review: 0.3 Findings. Independent Instance: 3.7. Human Baseline: 4.1.](https://certified-architect.com/cca-f-study-guide/domain4/k4-6-1-multi-pass-review-architecture): Multi-pass review architecture
- [K4.6.2 Even Design Goals from the Generation Prompt Suppress Review Findings](https://certified-architect.com/cca-f-study-guide/domain4/k4-6-2-per-file-cross-file-passes): Per-file + cross-file passes
- [K4.6.3 Single-Pass Review at 13+ Files: 43% Detection. Multi-Pass: 86%.](https://certified-architect.com/cca-f-study-guide/domain4/k4-6-3-independent-instances-eliminate-bias): Independent instances eliminate bias
- [S4.6.1 Calibrate Confidence Per Category — "HIGH" Means 92% for Style but 70% for Security](https://certified-architect.com/cca-f-study-guide/domain4/s4-6-1-multi-instance-review-pipeline): Multi-instance review pipeline

### Context Management & Reliability

- [K5.1.1 "$127.50 Refund" Became "Customer Requested a Refund" — The Agent Processed $50](https://certified-architect.com/cca-f-study-guide/domain5/k5-1-1-progressive-summarization-case-facts): Progressive summarization & Case Facts
- [K5.1.2 Sources 1-2: 96%. Sources 5-6: 52%. Sources 9-10: 94%. The U-Shaped Attention Curve.](https://certified-architect.com/cca-f-study-guide/domain5/k5-1-2-lost-in-the-middle): Lost-in-the-middle effect
- [K5.1.3 40 Fields Returned, 5 Needed — Context Full After 4 Calls Instead of 11](https://certified-architect.com/cca-f-study-guide/domain5/k5-1-3-tool-output-bloat-filtering): Tool output bloat and filtering
- [K5.1.4 The API Does Not Remember Your Conversation — You Must Send It Every Time](https://certified-architect.com/cca-f-study-guide/domain5/k5-1-4-stateless-api-history-management): Stateless API & history management
- [S5.1.1 Turns 1-10: 96% Accuracy. Turns 31+: 58%. No Persistent Facts Block.](https://certified-architect.com/cca-f-study-guide/domain5/s5-1-1-case-facts-issue-tracker): Case Facts & issue tracker pattern
- [S5.1.2 28% of Issues Dropped Without a Tracker. 3% With One.](https://certified-architect.com/cca-f-study-guide/domain5/s5-1-2-multi-issue-tracking): Multi-issue tracking
- [S5.1.3 2,800 Tokens of Reasoning Chain → 280 Tokens of Structured Facts. Same Findings, 10x Less Context.](https://certified-architect.com/cca-f-study-guide/domain5/s5-1-3-sub-agent-structured-output): Sub-agent structured output efficiency
- [K5.2.1 Sentiment-Based Escalation: 40% Volume, 30% Needed Human. Replace It.](https://certified-architect.com/cca-f-study-guide/domain5/k5-2-1-escalation-trigger-design): Escalation trigger design
- [K5.2.2 "I Want a Person" → Attempt Resolution First → CSAT 2.1. Immediate Escalation → 3.8.](https://certified-architect.com/cca-f-study-guide/domain5/k5-2-2-immediate-escalation-human-request): Immediate escalation on human request
- [K5.2.3 High Confidence (0.9+): 12% Errors. Low Confidence (<0.5): 68% Correct. The Signal Is Broken.](https://certified-architect.com/cca-f-study-guide/domain5/k5-2-3-confidence-scores-unreliable): Confidence scores unreliable
- [K5.2.4 Auto-Select "Most Recent": 27% Wrong Customer. Ask for Email: 2%.](https://certified-architect.com/cca-f-study-guide/domain5/k5-2-4-multi-match-disambiguation): Multi-match disambiguation
- [S5.2.1 Policy Silent on Competitor Matching → Agent Decides → 52% Approve, 48% Deny. Same Request.](https://certified-architect.com/cca-f-study-guide/domain5/s5-2-1-policy-gap-detection-escalation): Policy gap detection & escalation
- [K5.3.1 Generic "Failed" → 18% Recovery. Structured Error → 71%.](https://certified-architect.com/cca-f-study-guide/domain5/k5-3-1-structured-error-context): Structured error context for recovery
- [K5.3.2 "Database Error" — Orchestrator Retried 5 Times. The Database Was Permanently Decommissioned.](https://certified-architect.com/cca-f-study-guide/domain5/k5-3-2-generic-error-anti-pattern): Generic error status anti-pattern
- [K5.3.3 Two Anti-Patterns That Compound: 35% of Reports Have Hidden Gaps, 25% of Queries Are Killed](https://certified-architect.com/cca-f-study-guide/domain5/k5-3-3-silent-swallow-termination-anti-patterns): Silent swallow + termination anti-patterns
- [K5.3.4 "No Peer-Reviewed Studies Exist" — Actually, 47 Papers Were Found After the Outage Ended](https://certified-architect.com/cca-f-study-guide/domain5/k5-3-4-access-failure-vs-valid-empty): Access failure vs valid empty
- [S5.3.1 "AI Has Minimal Impact on Performing Arts" — Actually, the Search Just Timed Out](https://certified-architect.com/cca-f-study-guide/domain5/s5-3-1-coverage-annotation-synthesis): Coverage annotation in synthesis
- [K5.4.1 Minute 0: "src/auth/jwt.ts:12 → verifyJWT()". Minute 45: "Typical JWT Validation Pattern."](https://certified-architect.com/cca-f-study-guide/domain5/k5-4-1-context-degradation-over-time): Context degradation over time
- [K5.4.2 The Scratchpad: Writing Findings Down Before the Context Forgets Them](https://certified-architect.com/cca-f-study-guide/domain5/k5-4-2-scratchpad-persistent-findings): Scratchpad for persistent findings
- [K5.4.3 Sub-Agent Delegation: Let Someone Else Hold the Data](https://certified-architect.com/cca-f-study-guide/domain5/k5-4-3-sub-agent-delegation-isolation): Sub-agent delegation for context isolation
- [K5.4.4 Crash Recovery: A Manifest File So You Don't Start Over](https://certified-architect.com/cca-f-study-guide/domain5/k5-4-4-crash-recovery-manifest): Crash recovery with manifest
- [S5.4.1 Save, Compact, Restore: The Three-Step Workflow for /compact](https://certified-architect.com/cca-f-study-guide/domain5/s5-4-1-compact-save-restore-workflow): /compact save → compact → restore
- [K5.5.1 96% Accuracy, 40% More Escalations: Why Aggregate Metrics Lie](https://certified-architect.com/cca-f-study-guide/domain5/k5-5-1-aggregate-accuracy-masks-failures): Aggregate accuracy masks failures
- [K5.5.2 Stratified Sampling: Why Random Monitoring Misses the Problems That Matter](https://certified-architect.com/cca-f-study-guide/domain5/k5-5-2-stratified-sampling-monitoring): Stratified sampling for monitoring
- [K5.5.3 Field-Level Confidence: Review the Uncertain Fields, Not the Entire Document](https://certified-architect.com/cca-f-study-guide/domain5/k5-5-3-field-level-confidence-routing): Field-level confidence routing
- [K5.5.4 Confidence Calibration: The Model Says 0.9 — What Does That Actually Mean?](https://certified-architect.com/cca-f-study-guide/domain5/k5-5-4-confidence-threshold-calibration): Confidence threshold calibration
- [K5.6.1 Claim-Source Mapping: Every Fact Needs a Return Address](https://certified-architect.com/cca-f-study-guide/domain5/k5-6-1-claim-source-mapping-provenance): Claim-source mapping provenance
- [K5.6.2 Conflicting Data: Present Both, Fabricate Neither](https://certified-architect.com/cca-f-study-guide/domain5/k5-6-2-conflicting-data-preserve-both): Conflicting data: preserve both
- [K5.6.3 Temporal Metadata: Not Every Difference Is a Contradiction](https://certified-architect.com/cca-f-study-guide/domain5/k5-6-3-temporal-metadata-false-contradictions): Temporal metadata prevents false contradictions
- [K5.6.4 Established vs Contested: Structure Reports by Evidence Strength](https://certified-architect.com/cca-f-study-guide/domain5/k5-6-4-well-established-vs-contested): Well-established vs contested structure
- [K5.6.5 Content-Type-to-Format Matching: Tables for Numbers, Prose for Analysis](https://certified-architect.com/cca-f-study-guide/domain5/k5-6-5-content-type-format-matching): Content-type-to-format matching

## Machine-readable indexes

- [Article Index (English)](https://certified-architect.com/article-index.json): JSON map of all study articles by domain and knowledge point.
- [Article Index (Chinese)](https://certified-architect.com/article-index.zh.json): same structure, Simplified Chinese.
- [Question Pool Index](https://certified-architect.com/question-pools/index.json): per-domain question counts and mock exam sampling config.
- [Sitemap](https://certified-architect.com/sitemap-index.xml)

## Official resources (not part of this site)

- [Anthropic Academy](https://anthropic.skilljar.com/): official courses and CCA-F registration.
- [Pearson VUE — Claude Certification Program](https://www.pearsonvue.com/us/en/anthropic.html): exam delivery, scheduling.

## Localization

Simplified Chinese versions live under `/zh/` with the same paths (e.g. https://certified-architect.com/zh/cca-f-study-guide, https://certified-architect.com/zh/blog). English and Chinese articles share slugs and knowledge-point IDs.

## Legacy URLs

`/knowledge/*`, `/practice`, `/practice/mock-exam`, `/reference` (and `/zh/` equivalents) permanently redirect (301) to the URLs above.