Ace the Claude Certified Architect exam Adaptive certification prep powered by the Model Context Protocol. 390 questions. Guided capstone build. 30 concept handouts. 6 reference projects. Practice exams. Spaced repetition. Zero sycophancy. Quick Start • Features • How It Works • Study Modes • Capstone Build • Reference Projects • Concept Handouts • Exam Domains • Architecture Architect Cert is a f
Add this skill
npx mdskills install connectry-io/connectrylab-architect-cert-mcpComprehensive certification prep with 390 questions, spaced repetition, and zero sycophancy
1<p align="center">2 <br />3 <picture>4 <source media="(prefers-color-scheme: dark)" srcset="https://raw.githubusercontent.com/Connectry-io/connectrylab-architect-cert-mcp/master/.github/assets/logo-dark.svg">5 <source media="(prefers-color-scheme: light)" srcset="https://raw.githubusercontent.com/Connectry-io/connectrylab-architect-cert-mcp/master/.github/assets/logo-light.svg">6 <img alt="Architect Cert" src="https://raw.githubusercontent.com/Connectry-io/connectrylab-architect-cert-mcp/master/.github/assets/logo-dark.svg" width="420">7 </picture>8 <br />9</p>1011<h3 align="center">12 Ace the Claude Certified Architect exam13</h3>1415<p align="center">16 Adaptive certification prep powered by the Model Context Protocol.<br />17 390 questions. Guided capstone build. 30 concept handouts. 6 reference projects. Practice exams. Spaced repetition. Zero sycophancy.18</p>1920<p align="center">21 <a href="https://www.npmjs.com/package/connectry-architect-mcp"><img src="https://img.shields.io/npm/v/connectry-architect-mcp?style=flat&colorA=18181B&colorB=E8784A" alt="npm version"></a>22 <a href="https://www.npmjs.com/package/connectry-architect-mcp"><img src="https://img.shields.io/npm/dm/connectry-architect-mcp?style=flat&colorA=18181B&colorB=E8784A" alt="npm downloads"></a>23 <a href="https://github.com/Connectry-io/connectrylab-architect-cert-mcp"><img src="https://img.shields.io/github/stars/Connectry-io/connectrylab-architect-cert-mcp?style=flat&colorA=18181B&colorB=E8784A" alt="GitHub stars"></a>24 <a href="https://github.com/Connectry-io/connectrylab-architect-cert-mcp/blob/master/LICENSE"><img src="https://img.shields.io/github/license/Connectry-io/connectrylab-architect-cert-mcp?style=flat&colorA=18181B&colorB=E8784A" alt="License"></a>25</p>2627<p align="center">28 <a href="#-quick-start">Quick Start</a> •29 <a href="#-features">Features</a> •30 <a href="#-how-it-works">How It Works</a> •31 <a href="#-study-modes">Study Modes</a> •32 <a href="#-guided-capstone-build">Capstone Build</a> •33 <a href="#-reference-projects">Reference Projects</a> •34 <a href="#-concept-handouts">Concept Handouts</a> •35 <a href="#-exam-domains">Exam Domains</a> •36 <a href="#-tools">Tools</a> •37 <a href="#-architecture">Architecture</a>38</p>3940---4142## What is Architect Cert?4344Architect Cert is a free, open-source [MCP](https://modelcontextprotocol.io/) server that turns Claude into your personal certification tutor for the **Claude Certified Architect — Foundations** exam. No courses, no slides, no video lectures — just ask Claude and study.4546It ships with:47- **390 scenario-based questions** across all 5 exam domains and 30 task statements48- **Guided capstone build** — shape your own project, then build it step-by-step while learning every task statement hands-on49- **30 concept handouts** — one per task statement, with code examples and common mistakes50- **6 reference projects** — runnable TypeScript codebases demonstrating each domain in practice51- **Practice exams** — 60-question weighted exams with history tracking and improvement trends52- **Interactive follow-ups** — wrong answer? Dive into code examples, concept lessons, handouts, or reference projects before moving on53- **PDF generation** — branded handout PDFs with the Architect Cert logo for offline study54- **Spaced repetition** — SM-2 algorithm schedules reviews at optimal intervals55- **Deterministic grading** — pure function grading, no LLM judgment, zero sycophancy5657Everything runs locally. No cloud, no accounts, no telemetry.5859<br />6061## Quick Start6263### 1. Install6465```bash66npm install -g connectry-architect-mcp67```6869### 2. Configure Your MCP Client7071<details>72<summary><b>Claude Code</b> (VS Code / Cursor / Terminal)</summary>7374Add to `.mcp.json` in your project or `~/.claude.json` globally:7576```json77{78 "mcpServers": {79 "connectry-architect": {80 "command": "connectry-architect-mcp"81 }82 }83}84```8586Then restart Claude Code. The server starts automatically when Claude loads.8788</details>8990<details>91<summary><b>Claude Desktop</b> — macOS</summary>9293Add to `~/Library/Application Support/Claude/claude_desktop_config.json`:9495```json96{97 "mcpServers": {98 "connectry-architect": {99 "command": "connectry-architect-mcp"100 }101 }102}103```104105Restart Claude Desktop. You'll see the MCP tools icon appear in the chat input.106107</details>108109<details>110<summary><b>Claude Desktop</b> — Windows</summary>111112Add to `%APPDATA%\Claude\claude_desktop_config.json`:113114```json115{116 "mcpServers": {117 "connectry-architect": {118 "command": "connectry-architect-mcp"119 }120 }121}122```123124Restart Claude Desktop. You'll see the MCP tools icon appear in the chat input.125126</details>127128<details>129<summary><b>Any MCP-compatible client</b></summary>130131Architect Cert works with any client that supports the [Model Context Protocol](https://modelcontextprotocol.io/). Configure it as a stdio server:132133- **Command:** `connectry-architect-mcp`134- **Arguments:** none135- **Transport:** stdio136137The server exposes 17 tools, 8 prompts, and 3 resource types.138139</details>140141### 3. Start Studying142143Restart your MCP client and start chatting:144145| What you want | What to ask Claude |146|---------------|-------------------|147| Find your weak areas | *"Start an assessment to figure out where I stand"* |148| Practice questions | *"Give me a practice question"* |149| Focus on a domain | *"Give me a question about agentic architecture"* |150| Learn a concept first | *"Teach me about task 2.3 — tool provisioning"* |151| Build your own capstone | *"I want to start a guided capstone build"* |152| Take a practice exam | *"I want to take a practice exam"* |153| Check your progress | *"Show my study progress"* |154| Get a study plan | *"What should I study next?"* |155| Explore a reference project | *"Show me a reference project for domain 1"* |156| Generate PDF handouts | Run `npm run generate:pdfs` in the project directory |157| Reset and start over | *"Reset my progress"* |158159<br />160161## Features162163<table>164<tr>165<td width="50%">166167### Adaptive Learning168SM-2 spaced repetition resurfaces weak areas at optimal intervals. Questions you get wrong come back sooner — questions you master fade away.169170### 390 Exam Questions17113 scenario-based questions per task statement across all 30 topics. Easy, medium, and hard difficulties with balanced answer distributions across A/B/C/D.172173### Deterministic Grading174Pure function grading — no LLM judgment calls. Your answer is checked against a verified key. Right is right, wrong is wrong. Every wrong answer includes a specific explanation.175176</td>177<td width="50%">178179### Zero Sycophancy180Claude won't sugarcoat wrong answers. Anti-sycophancy rules are enforced at the protocol level. Wrong means wrong — no "you were on the right track."181182### Progress Tracking183Persistent SQLite database tracks every answer, mastery level, and review schedule. Pick up exactly where you left off across sessions, devices, and MCP clients.184185### Smart Question Selection186Three-priority algorithm: overdue reviews first, then weak areas, then new material. You always work on what matters most for exam readiness.187188</td>189</tr>190<tr>191<td width="50%">192193### Interactive Follow-Ups194After every answer, you get follow-up options. Got it wrong? Choose to see a code example, read the concept lesson, open the handout, or explore the relevant reference project — then jump back to your quiz.195196### Guided Capstone Build197Shape your own project idea, then build it file-by-file across 18 steps. Each step: quiz on the relevant task statements, Claude generates themed code, then a walkthrough explains how every section maps to exam concepts. Learn by doing.198199</td>200<td width="50%">201202### 30 Concept Handouts203Every task statement has a structured handout: concept explanation, TypeScript code example, common mistakes, and documentation references. Available as markdown in Claude or as branded PDFs for offline study.204205### 6 Reference Projects206Runnable TypeScript codebases that demonstrate each domain in practice. A capstone project ties all 5 domains together. Each file maps to specific task statements so you can see concepts in real code.207208### Practice Exams209Full 60-question exams weighted by domain — just like the real certification. Fresh questions each time (no repetition from your last attempt). Scored out of 1000, passing at 720. All attempts saved with per-domain breakdowns.210211### PDF Handouts212Generate branded PDFs for all 30 handouts with `npm run generate:pdfs`. Each PDF includes the Architect Cert logo, domain label, and clean formatting for printing or tablet reading.213214</td>215</tr>216</table>217218<br />219220## How It Works221222```223┌──────────────────────────────────────────────────────────────────┐224│ YOUR MACHINE │225│ │226│ ┌──────────────┐ ┌──────────────────────────┐ │227│ │ Claude Desktop│ │ Architect Cert MCP │ │228│ │ Claude Code │◄─────►│ │ │229│ │ Any MCP client│ stdio │ 17 tools │ │230│ └──────────────┘ │ 8 prompts │ │231│ │ 3 resource types │ │232│ └──────────┬───────────────┘ │233│ │ │234│ ┌─────────────────┼─────────────────┐ │235│ │ │ │ │236│ ~/.connectry- 390 questions 6 reference │237│ architect/ 30 handouts projects │238│ progress.db (bundled JSON/MD) (bundled TS) │239│ │240└──────────────────────────────────────────────────────────────────┘241```242243**The study loop:**2442451. **Assessment** — 15 diagnostic questions baseline your knowledge across all 5 domains2462. **Path assignment** — Score < 60% gets the beginner-friendly track; score >= 60% gets the exam-weighted track2473. **Learn first** — Read the concept handout for a topic before tackling questions (or skip straight to practice)2484. **Smart selection** — Each question chosen by priority: overdue reviews > weak areas > new material2495. **Deterministic grading** — Pure function checks your answer against the verified key2506. **Interactive follow-up** — Wrong answer? Dive into code examples, concept lessons, or reference projects2517. **Spaced repetition** — SM-2 algorithm schedules reviews: 1 day > 3 days > 6 days > ...2528. **Practice exams** — Take full 60-question exams when you're ready, track improvement over time253254**Data storage:**255256- Progress is stored locally at `~/.connectry-architect/progress.db` (SQLite, WAL mode)257- Your user config lives at `~/.connectry-architect/config.json` (auto-created on first run)258- No cloud, no accounts, no telemetry — everything stays on your machine259260<br />261262## Study Modes263264### Initial Assessment265266Start with 15 questions across all domains. Based on your score:267268- **< 60% accuracy** — **Beginner-Friendly Path**: Starts with fundamentals, builds up gradually through each domain. Focuses on easy and medium questions first.269- **>= 60% accuracy** — **Exam-Weighted Path**: Focuses on high-weight domains first (D1 at 27%, D3 & D4 at 20% each). Targets weak areas aggressively.270271### Learn-First Mode272273Before tackling questions on any topic, you can read the concept handout first:274275```276You: "Teach me about task 1.5 — tool-use hooks"277278Claude: [loads the concept handout with key concepts, code examples,279 common mistakes, and documentation references]280281 Ready to test your understanding? I have 13 questions282 on this topic ranging from easy to hard.283```284285The server tracks which handouts you've viewed. You can always come back to them later.286287### Interactive Follow-Ups288289After every answer, you get contextual follow-up options:290291**After a wrong answer:**292```293The correct answer is B. [explanation]294295 > Got it, next question296 > Explain with a code example297 > Show me the concept lesson298 > Show me the handout299 > Show me in the reference project300```301302**After a correct answer:**303```304Correct!305306 > Next question307 > Explain why the others are wrong308```309310Each option dives deeper into the concept — then brings you right back to your quiz. This means you never have to leave your study flow to look something up.311312### Adaptive Practice313314Every question is selected by a three-priority algorithm:3153161. **Overdue reviews** — Spaced repetition items due for review today3172. **Weak areas** — Topics where your mastery is below 50%3183. **New material** — Fresh questions from your recommended domain319320### Practice Exams321322Full 60-question exams that simulate the real certification:323324| Detail | Value |325|--------|-------|326| Total questions | 60 |327| D1: Agentic Architecture | 16 questions (27%) |328| D2: Tool Design & MCP | 11 questions (18%) |329| D3: Claude Code Config | 12 questions (20%) |330| D4: Prompt Engineering | 12 questions (20%) |331| D5: Context & Reliability | 9 questions (15%) |332| Scoring | 0–1000, passing at 720 |333| Question selection | Fresh set each time — avoids repeating your most recent attempt |334335All attempts are saved with per-domain score breakdowns and improvement trends.336337### Mastery Levels338339Each of the 30 task statements has an independent mastery level:340341| Level | Criteria | What it means |342|-------|----------|---------------|343| **Unassessed** | No attempts yet | You haven't seen questions on this topic |344| **Weak** | < 50% accuracy | Needs significant study — questions resurface frequently |345| **Developing** | 50–69% accuracy | Making progress — keep practicing |346| **Strong** | 70–89% accuracy | Good understanding — review intervals are longer |347| **Mastered** | >= 90% accuracy, 5+ attempts, 3+ consecutive correct | Exam-ready — rare reviews |348349### Spaced Repetition (SM-2)350351The [SM-2 algorithm](https://en.wikipedia.org/wiki/SuperMemo#Description_of_SM-2_algorithm) schedules review intervals:352353- **First review:** 1 day after answering354- **Second review:** 3 days after first review355- **Subsequent reviews:** Previous interval x ease factor (starts at 2.5)356- **Wrong answer:** Interval resets, ease factor decreases by 0.2 (floor: 1.3)357- **Correct answer:** Ease factor increases by 0.1358359Difficult questions come back often. Easy ones space out to weeks or months.360361<br />362363## Guided Capstone Build364365The most hands-on way to learn — build your own project from scratch while covering all 30 task statements. Instead of just answering questions, you architect a real system themed to your own idea.366367### How it works368369The capstone build has three phases:370371**Phase 1 — Project Shaping**372373You describe a project idea (e.g., "a multi-agent code review system"). Claude analyzes your idea against all 30 architectural criteria and identifies gaps. You refine together until every task statement is covered.374375```376You: "I want to start a guided capstone build"377378Claude: [presents the 30 criteria across all 5 domains]379 Describe your project idea and I'll analyze coverage.380381You: "A multi-agent code review system that analyzes PRs"382383Claude: Your idea naturally covers 24/30 criteria. To cover the384 remaining 6, I'd suggest adding: [specific suggestions385 mapped to task statements]386```387388**Phase 2 — Interleaved Build (18 steps)**389390Each step follows the same pattern:3913921. **Quiz** — 2-3 questions on the task statements you're about to build3932. **Build** — Claude generates the file's code, themed to your project3943. **Walkthrough** — Line-by-line explanation mapping code to task statements395396The 18 steps build incrementally:397398| Steps | What you build | Task Statements |399|-------|---------------|-----------------|400| 1-2 | Project config (CLAUDE.md, package.json) | 3.1-3.4 |401| 3-5 | MCP server, tools, error handling | 2.1-2.5 |402| 6-10 | Agentic loop, subagents, hooks, workflows, sessions | 1.1-1.7 |403| 11-13 | Prompts: system, extraction, batch processing | 4.1-4.6 |404| 14-18 | Context: preservation, triggers, propagation, scratchpad, confidence | 5.1-5.6 |405406Every quiz answer feeds into the same spaced repetition and mastery tracking as regular practice.407408**Phase 3 — Final Review**409410After step 18, you get a complete coverage map: all 30 task statements, where each is demonstrated in your project, and your quiz performance per domain. Weak areas are flagged for further study.411412### Capstone build tools413414| Tool | What it does |415|------|-------------|416| `start_capstone_build` | See the 30 criteria, describe your theme, refine until coverage is complete |417| `capstone_build_step` | Drive the build: confirm, quiz, build, next, status, or abandon |418| `capstone_build_status` | Check your progress — current step, criteria coverage, quiz performance |419420### How it connects to everything else421422- Quiz answers during the build use the same `submit_answer` grading and SM-2 scheduling423- After any quiz question, you can use the same follow-up options (code example, concept lesson, handout, reference project)424- The reference projects show how the capstone structure looks when complete425- Progress persists across sessions — pick up where you left off426427<br />428429## Reference Projects430431Architect Cert includes **6 complete reference projects** — runnable TypeScript codebases that demonstrate certification concepts in real code. Every file has a header comment mapping it to specific task statements.432433| Project | Focus | Files | What You'll See |434|---------|-------|-------|-----------------|435| **Capstone** | All 5 domains | 24 | Full multi-agent support system with MCP server, coordinator, subagents, prompt engineering, context management, and hooks |436| **D1 — Agentic Loop** | Domain 1 | 10 | Multi-agent research coordinator with agentic loops, subagent spawning, hooks, session management, and task decomposition |437| **D2 — Tool Design** | Domain 2 | 12 | MCP server with split tools, structured errors, agent-scoped tool distribution, resources, and built-in tool patterns |438| **D3 — Claude Code Config** | Domain 3 | 14 | Complete config reference: CLAUDE.md hierarchy, slash commands, path rules, CI/CD workflows — not runnable code, but a real config layout |439| **D4 — Prompt Engineering** | Domain 4 | 11 | Data extraction pipeline with explicit criteria, few-shot, structured output, validation-retry, batch processing, and multi-pass review |440| **D5 — Context Manager** | Domain 5 | 14 | Long-session patterns: context preservation, scratchpad, subagent delegation, escalation, error propagation, confidence calibration, provenance |441442### How to access them443444Ask Claude:445446```447You: "Show me a reference project for domain 1"448449Claude: [calls scaffold_project, returns the project README,450 file listing, and architecture walkthrough]451```452453Or browse them directly in the repo under `projects/`.454455### How they connect to the study flow456457When you get a question wrong, one of the follow-up options is **"Show me in the reference project"** — this takes you straight to the relevant domain project so you can see the concept implemented in real code. Then you jump back to your quiz.458459<br />460461## Concept Handouts462463Every task statement has a **concept handout** — a structured study document (~500-800 words) that covers:464465- **Concept** — The core idea, mental model, and when/why to use it466- **Code Example** — Realistic TypeScript demonstrating the pattern467- **Common Mistakes** — The 3-5 most frequent errors (which map to exam wrong answers)468- **References** — Links to Anthropic's official documentation469470### Reading handouts in Claude471472```473You: "Show me the handout for task 2.3"474475Claude: [loads the full handout with concept, code, mistakes, references]476```477478### Generating PDF handouts479480You can generate branded PDFs for all 30 handouts for offline study:481482```bash483cd connectrylab-architect-cert-mcp484npm run generate:pdfs485```486487This creates 30 PDFs in `generated/handouts/` with:488- Architect Cert logo and domain label in the header489- Clean formatting with syntax-highlighted code blocks490- "Connectry LABS — Claude Certified Architect Exam Prep — Free & Open Source" footer491492<br />493494## Exam Domains495496The Claude Certified Architect — Foundations exam covers 5 domains:497498| # | Domain | Weight | Tasks | Questions |499|---|--------|--------|-------|-----------|500| 1 | Agentic Architecture & Orchestration | 27% | 7 | 91 |501| 2 | Tool Design & MCP Integration | 18% | 5 | 65 |502| 3 | Claude Code Configuration & Workflows | 20% | 6 | 78 |503| 4 | Prompt Engineering & Structured Output | 20% | 6 | 78 |504| 5 | Context Management & Reliability | 15% | 6 | 78 |505| | **Total** | **100%** | **30** | **390** |506507### 30 Task Statements508509<details>510<summary><b>Domain 1 — Agentic Architecture & Orchestration</b> (7 tasks, 91 questions)</summary>511512| Task | Description |513|------|-------------|514| 1.1 | Design and implement agentic loops for autonomous task execution |515| 1.2 | Orchestrate multi-agent systems with coordinator-subagent patterns |516| 1.3 | Configure subagent invocation, context passing, and spawning |517| 1.4 | Implement multi-step workflows with enforcement and handoff patterns |518| 1.5 | Apply Agent SDK hooks for tool call interception and data normalization |519| 1.6 | Design task decomposition strategies for complex workflows |520| 1.7 | Manage session state, resumption, and forking |521522</details>523524<details>525<summary><b>Domain 2 — Tool Design & MCP Integration</b> (5 tasks, 65 questions)</summary>526527| Task | Description |528|------|-------------|529| 2.1 | Design effective tool interfaces with clear descriptions and boundaries |530| 2.2 | Implement structured error responses for MCP tools |531| 2.3 | Distribute tools appropriately across agents and configure tool choice |532| 2.4 | Integrate MCP servers into Claude Code and agent workflows |533| 2.5 | Select and apply built-in tools effectively |534535</details>536537<details>538<summary><b>Domain 3 — Claude Code Configuration & Workflows</b> (6 tasks, 78 questions)</summary>539540| Task | Description |541|------|-------------|542| 3.1 | Configure CLAUDE.md files with appropriate hierarchy and scoping |543| 3.2 | Create and configure custom slash commands and skills |544| 3.3 | Apply path-specific rules for conditional convention loading |545| 3.4 | Determine when to use plan mode vs direct execution |546| 3.5 | Apply iterative refinement techniques for progressive improvement |547| 3.6 | Integrate Claude Code into CI/CD pipelines |548549</details>550551<details>552<summary><b>Domain 4 — Prompt Engineering & Structured Output</b> (6 tasks, 78 questions)</summary>553554| Task | Description |555|------|-------------|556| 4.1 | Design prompts with explicit criteria to improve precision |557| 4.2 | Apply few-shot prompting to improve output consistency |558| 4.3 | Enforce structured output using tool use and JSON schemas |559| 4.4 | Implement validation, retry, and feedback loops |560| 4.5 | Design efficient batch processing strategies |561| 4.6 | Design multi-instance and multi-pass review architectures |562563</details>564565<details>566<summary><b>Domain 5 — Context Management & Reliability</b> (6 tasks, 78 questions)</summary>567568| Task | Description |569|------|-------------|570| 5.1 | Manage conversation context to preserve critical information |571| 5.2 | Design effective escalation and ambiguity resolution patterns |572| 5.3 | Implement error propagation strategies across multi-agent systems |573| 5.4 | Manage context effectively in large codebase exploration |574| 5.5 | Design human review workflows and confidence calibration |575| 5.6 | Preserve information provenance and handle uncertainty in synthesis |576577</details>578579<br />580581## Tools582583Architect Cert provides **17 MCP tools** that Claude uses to deliver the study experience:584585| Tool | Description |586|------|-------------|587| `start_assessment` | Begin with 15 diagnostic questions to determine your learning path |588| `get_practice_question` | Get the next adaptive question (reviews > weak areas > new material) |589| `submit_answer` | Grade your answer deterministically — includes interactive follow-up options |590| `follow_up` | Handle post-answer actions: code examples, concept lessons, handouts, reference projects |591| `get_progress` | View overall study progress with mastery percentages per domain |592| `get_curriculum` | Browse all 5 domains and 30 task statements with current mastery levels |593| `get_section_details` | Deep dive into a specific task statement with concept lesson |594| `get_weak_areas` | Identify topics that need the most work, ranked by weakness |595| `get_study_plan` | Get personalized recommendations based on performance and exam weights |596| `start_practice_exam` | Take a full 60-question practice exam simulating the real certification |597| `submit_exam_answer` | Submit and grade answers during a practice exam |598| `get_exam_history` | View all past exam attempts with scores, trends, and per-domain comparison |599| `scaffold_project` | Access reference projects for hands-on practice with real code |600| `start_capstone_build` | Start a guided capstone build — shape your project and validate criteria coverage |601| `capstone_build_step` | Drive the capstone build: confirm, quiz, build, next, status, or abandon |602| `capstone_build_status` | Check capstone build progress — current step, coverage, quiz performance |603| `reset_progress` | Start over — requires explicit confirmation to prevent accidents |604605The server also registers **8 interactive prompts** and **3 resource types** (concept handouts, reference projects, exam overview).606607<br />608609## Architecture610611```612Claude (UI) <-> MCP Server (stdio) <-> Core Engine <-> SQLite613```614615| Component | Technology | Purpose |616|-----------|-----------|---------|617| MCP Server | `@modelcontextprotocol/sdk` v1 | Registers tools, prompts, resources over stdio |618| Grading Engine | Pure TypeScript functions | Deterministic answer verification |619| Spaced Repetition | SM-2 algorithm | Optimal review scheduling |620| Follow-Up System | State-driven tool chain | Interactive post-answer detours |621| Capstone Build Engine | 18-step interleaved builder | Guided learn-build-explain flow with LLM validation |622| Question Bank | 390 bundled JSON questions | Scenario-based, verified against docs |623| Concept Handouts | 30 bundled markdown files | Structured study materials per task statement |624| Reference Projects | 6 bundled TypeScript projects | Runnable code demonstrating each domain |625| PDF Generator | `pdfkit` | Branded handout PDFs for offline study |626| Progress Store | `better-sqlite3` (WAL mode) | Persistent mastery, answers, schedules |627628### Anti-Sycophancy Design629630This server enforces honest grading at the protocol level — not just in prompts:6316321. **Deterministic grading** — `gradeAnswer()` is a pure function. No LLM is involved in judging correctness.6332. **Tool-level enforcement** — The `submit_answer` tool description instructs Claude to relay results verbatim.6343. **No partial credit** — Multiple choice, one correct answer. No "you were on the right track."6354. **Wrong answer explanations** — Every incorrect option has a specific `whyWrongMap` entry explaining the misconception.6365. **System prompt rules** — Five anti-sycophancy directives prevent Claude from softening incorrect results.637638<br />639640## Question Bank Details641642| Metric | Value |643|--------|-------|644| Total questions | 390 |645| Domains covered | 5 |646| Task statements covered | 30 |647| Questions per task statement | 13 |648| Difficulty distribution | ~4 easy, 5 medium, ~4 hard per task |649| Answer key balance | Distributed across A/B/C/D |650| Question format | Scenario-based multiple choice |651| Each question includes | Scenario, question, 4 options, explanation, why-wrong-map, references |652| Source material | Anthropic official documentation |653654<br />655656## Contributing657658We welcome contributions! Here's how to get started:659660```bash661# Clone the repo662git clone https://github.com/Connectry-io/connectrylab-architect-cert-mcp.git663cd connectrylab-architect-cert-mcp664665# Install dependencies666npm install667668# Build669npm run build670671# Run tests (30 tests across 6 test files)672npm test673674# Generate PDF handouts675npm run generate:pdfs676677# Run locally678node dist/index.js679```680681### Project Structure682683```684src/685├── index.ts # MCP server entry point686├── config.ts # User config management687├── types.ts # All TypeScript interfaces688├── data/689│ ├── loader.ts # Lazy-cached data loading690│ ├── curriculum.json # 30 task statements691│ ├── questions/ # 390 questions (5 domain files)692│ ├── handouts/ # 30 concept handouts (markdown)693│ ├── criteria.ts # 30 task statement criteria for capstone validation694│ ├── build-steps.ts # 18 capstone build step definitions695│ └── system-prompt.ts # Anti-sycophancy rules696├── db/697│ ├── schema.ts # SQLite schema (9 tables)698│ ├── store.ts # Database initialization699│ ├── mastery.ts # Mastery level calculations700│ ├── answers.ts # Answer recording701│ ├── capstone.ts # Capstone build CRUD operations702│ └── exam-attempts.ts # Practice exam tracking703├── engine/704│ ├── grading.ts # Deterministic grading705│ ├── spaced-repetition.ts # SM-2 algorithm706│ ├── question-selector.ts # Priority-based selection707│ ├── exam-builder.ts # Practice exam generation708│ └── adaptive-path.ts # Learning path recommendations709├── tools/ # 17 MCP tool handlers710├── prompts/ # 8 MCP prompt definitions711└── resources/ # 3 MCP resource types712713projects/714├── capstone/ # All 5 domains — multi-agent support system715├── d1-agentic/ # Domain 1 — agentic loop research coordinator716├── d2-tools/ # Domain 2 — MCP server with tool patterns717├── d3-config/ # Domain 3 — Claude Code configuration layout718├── d4-prompts/ # Domain 4 — extraction & prompt engineering719└── d5-context/ # Domain 5 — context management & reliability720721scripts/722└── generate-pdfs.ts # PDF handout generator723```724725<br />726727## License728729MIT © [Connectry Labs](https://connectry.io/labs)730731<br />732733## Credits734735- [Anthropic](https://anthropic.com) — Claude & the Claude Certified Architect certification program736- [Model Context Protocol](https://modelcontextprotocol.io/) — The protocol that makes this possible737- [better-sqlite3](https://github.com/WiseLibs/better-sqlite3) — Fast, synchronous SQLite for Node.js738739<br />740741---742743<p align="center">744 <sub>Built with care by <a href="https://connectry.io/labs">Connectry Labs</a></sub>745</p>746747<p align="center">748 <a href="https://github.com/Connectry-io/connectrylab-architect-cert-mcp">GitHub</a> •749 <a href="https://www.npmjs.com/package/connectry-architect-mcp">npm</a> •750 <a href="https://connectry.io/labs">Website</a>751</p>752
Full transparency — inspect the skill content before installing.