# Connectry-io/connectrylab-architect-cert-mcp [Health: Active] **Category:** πŸŽ“ Education **Repository:** https://github.com/Connectry-io/connectrylab-architect-cert-mcp **GitHub Stars:** 33 **Views:** 2 **Installs:** 0 **Upvotes:** 0 **Directory Page:** https://allmcps.com/mcp/connectry-io-connectrylab-architect-cert-mcp ## Description Free certification prep for the Claude Certified Architect exam. 390 scenario-based questions, guided capstone build, 30 concept handouts, 6 reference projects, practice exams, interactive UI with clickable answer cards, glassmorphism progress dashboard, spaced repetition, and deterministic grading. ## Tools Capabilities this server exposes over MCP: - **start_assessment** β€” Begin with 15 diagnostic questions (lessons-first, one at a time) to determine your learning path - **get_practice_question** β€” Get the next adaptive question with clickable A/B/C/D buttons (reviews > weak areas > new material) - **submit_answer** β€” Grade your answer deterministically β€” presents interactive follow-up options - **follow_up** β€” Handle post-answer actions: code examples, concept lessons, handouts, reference projects - **get_section_details** β€” Deep dive into a specific task statement with full concept handout - **get_progress** β€” View overall study progress with mastery percentages per domain - **get_curriculum** β€” Browse all 5 domains and 30 task statements with current mastery levels - **get_weak_areas** β€” Identify topics that need the most work, ranked by weakness - **get_study_plan** β€” Get personalized recommendations with multi-select domain focus - **get_dashboard** β€” Open the visual progress dashboard in Claude Preview - **start_practice_exam** β€” Take a full 60-question practice exam simulating the real certification - **submit_exam_answer** β€” Submit and grade answers during a practice exam - **get_exam_history** β€” View all past exam attempts with scores, trends, and per-domain comparison - **start_capstone_build** β€” Start a guided capstone build β€” shape your project and validate criteria coverage - **capstone_build_step** β€” Drive the capstone build: confirm, quiz, build, next, status, or abandon - **capstone_build_status** β€” Check capstone build progress β€” current step, coverage, quiz performance - **scaffold_project** β€” Access reference projects for hands-on practice with real code - **reset_progress** β€” Start over β€” requires explicit confirmation to prevent accidents ## Claude Desktop Quick Installation Heuristic fallback β€” verify the package name and runner against the repository README before running it. Uses `npx` (confidence: low): ```json "mcpServers": { "connectrylab-architect-cert-mcp": { "command": "npx", "args": ["-y","connectry-io-connectrylab-architect-cert-mcp"] } } ``` ## Documentation & README


Architect Cert

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. Interactive dashboard. Spaced repetition. Zero sycophancy.

npm version npm downloads GitHub stars License Glama Score

Connectry Architect Cert MCP server

Quick Start β€’ Learning Path β€’ Interactive Experience β€’ Assessment β€’ Handouts β€’ Practice β€’ Capstone Build β€’ Practice Exams β€’ Dashboard β€’ Reference Projects β€’ Tools β€’ Architecture

--- ## What is Architect Cert? Architect Cert is a free, open-source [MCP](https://modelcontextprotocol.io/) server that turns Claude into your personal certification tutor for the **Claude Certified Architect β€” Foundations** exam. No courses, no slides, no video lectures β€” just ask Claude and study. It ships with: - **390 scenario-based questions** across all 5 exam domains and 30 task statements - **Interactive clickable UI** β€” answer questions with A/B/C/D buttons, select follow-ups, pick domains with checkboxes β€” all inside Claude - **Lessons-first assessment** β€” learn each domain's concepts before being tested on them - **Guided capstone build** β€” shape your own project, then build it step-by-step while learning every task statement hands-on - **30 concept handouts** β€” one per task statement, with code examples and common mistakes - **6 reference projects** β€” runnable TypeScript codebases demonstrating each domain in practice - **Practice exams** β€” 60-question weighted exams with history tracking and improvement trends - **Progress dashboard** β€” glassmorphism visual dashboard with mastery levels, exam history chart, activity timeline - **Visual progress tracking** β€” todo checklists track your progress through assessments, exams, and capstone builds in real time - **Interactive follow-ups** β€” wrong answer? Click to see code examples, concept lessons, handouts, or reference projects - **PDF generation** β€” branded handout PDFs with the Architect Cert logo for offline study - **Spaced repetition** β€” SM-2 algorithm schedules reviews at optimal intervals - **Deterministic grading** β€” pure function grading, no LLM judgment, zero sycophancy Everything runs locally. No cloud, no accounts, no telemetry.
## Quick Start ### 1. Install ```bash npm install -g connectry-architect-mcp ``` ### 2. Configure Your MCP Client
Claude Code β€” One-liner ```bash claude mcp add connectry-architect -- connectry-architect-mcp ``` That's it. Restart Claude Code and the server starts automatically.
Claude Code β€” Manual config Add to `.mcp.json` in your project or `~/.claude.json` globally: ```json { "mcpServers": { "connectry-architect": { "command": "connectry-architect-mcp" } } } ``` Restart Claude Code. The server starts automatically when Claude loads.
Claude Desktop β€” macOS Add to `~/Library/Application Support/Claude/claude_desktop_config.json`: ```json { "mcpServers": { "connectry-architect": { "command": "connectry-architect-mcp" } } } ``` > **Tip:** If you use `nvm`, you may need to specify the full path to the binary: > ```json > "command": "/Users/yourname/.nvm/versions/node/v22.20.0/bin/connectry-architect-mcp" > ``` Restart Claude Desktop. You'll see the MCP tools icon appear in the chat input.
Claude Desktop β€” Windows Add to `%APPDATA%\Claude\claude_desktop_config.json`: ```json { "mcpServers": { "connectry-architect": { "command": "connectry-architect-mcp" } } } ``` Restart Claude Desktop. You'll see the MCP tools icon appear in the chat input.
Any MCP-compatible client Architect Cert works with any client that supports the [Model Context Protocol](https://modelcontextprotocol.io/). Configure it as a stdio server: - **Command:** `connectry-architect-mcp` - **Arguments:** none - **Transport:** stdio The server exposes 18 tools, 8 prompts, and 3 resource types.
### 3. Start Studying Restart your MCP client and start chatting: | What you want | What to ask Claude | |---------------|-------------------| | Start from scratch | *"Start an assessment to figure out where I stand"* | | Practice questions | *"Give me a practice question"* | | Focus on a domain | *"Give me a question about agentic architecture"* | | Learn a concept first | *"Teach me about task 2.3 β€” tool provisioning"* | | Build your own capstone | *"I want to start a guided capstone build"* | | Take a practice exam | *"I want to take a practice exam"* | | Check your progress | *"Show my study progress"* | | Show the dashboard | *"Show my dashboard"* | | Get a study plan | *"What should I study next?"* | | Explore a reference project | *"Show me a reference project for domain 1"* | | Generate PDF handouts | Run `npm run generate:pdfs` in the project directory | | Reset and start over | *"Reset my progress"* |
## Learning Path Architect Cert is designed to follow a natural progression. Here's the recommended order: ``` 1. Assessment β†’ Baseline your knowledge across all 5 domains ↓ 2. Learn concepts β†’ Read handouts for weak domains before practicing ↓ 3. Adaptive practice β†’ Questions prioritized by weakness, reviews, new material ↓ 4. Follow-ups β†’ Dive into code examples, concept lessons, reference projects ↓ 5. Capstone build β†’ Build your own project covering all 30 task statements ↓ 6. Practice exams β†’ Simulate the real exam (60 questions, scored out of 1000) ↓ 7. Review & repeat β†’ Dashboard tracks mastery, spaced repetition handles scheduling ``` You can jump to any stage at any time β€” the system adapts. But the path above gives you the most structured experience.
## Interactive Experience Architect Cert uses Claude's built-in interactive UI to make studying feel natural. No typing answer letters β€” just click. ### Clickable Answer Selection Every question presents **clickable A/B/C/D buttons** directly in Claude. You tap your answer instead of typing it. If the question includes code in the scenario, a **code preview** appears alongside each option so you can reference it while deciding. ### Follow-Up Actions After every answer, clickable follow-up buttons appear: **After a wrong answer:** - Got it, next question - Explain with a code example - Show me the concept lesson - Show me the handout - Show me in the reference project **After a correct answer:** - Next question - Explain why the others are wrong ### Multi-Select Domain Picker When requesting a study plan, you can select multiple domains to focus on using **checkboxes** β€” no need to list them out by name. ### Visual Progress Checklists Long flows like the assessment (15 questions), practice exams (60 questions), and capstone build (18 steps) create a **visual todo checklist** that updates in real time as you progress. You always know where you are and what's left. ### Skip & Other Every question selection includes a **Skip** button to move on and an **Other** option to ask a free-form question. If you use "Other" to ask something, Claude answers your question and then **re-presents the same quiz question** β€” you never lose your place. "Skip" moves to the next question; skipped questions remain unanswered and will reappear later.
## Initial Assessment Start with a **15-question diagnostic** (3 per domain) that determines your learning path. The assessment works differently from raw practice: ### Lessons-First Flow When the assessment reaches a **new domain** for the first time, it pauses to show you the **concept handout** before asking questions. This means you learn the key ideas, see code examples, and understand common mistakes *before* being tested β€” no guessing blindly. The flow for each domain: 1. Concept handout is shown (one-time per domain) 2. 3 questions presented one at a time (easy, medium, hard) 3. Each answer graded immediately with explanation 4. Move to the next domain ### Path Assignment Based on your overall accuracy: | Score | Path | Description | |-------|------|-------------| | **< 60%** | Beginner-Friendly | Starts with fundamentals, builds up gradually. Focuses on easy and medium questions first. | | **>= 60%** | Exam-Weighted | Focuses on high-weight domains first (D1 at 27%, D3 & D4 at 20% each). Targets weak areas aggressively. | ### Progress Tracking The assessment creates a **15-item visual checklist** (Q1 through Q15, grouped by domain) that updates after each answer. You always see exactly how far you've progressed.
## Concept Handouts Every task statement has a **concept handout** β€” a structured study document (~500-800 words) that covers: - **Concept** β€” The core idea, mental model, and when/why to use it - **Code Example** β€” Realistic TypeScript demonstrating the pattern - **Common Mistakes** β€” The 3-5 most frequent errors (which map to exam wrong answers) - **References** β€” Links to Anthropic's official documentation ### Reading Handouts in Claude ``` You: "Show me the handout for task 2.3" Claude: [loads the full handout with concept, code, mistakes, references] ``` ### When Handouts Appear Automatically - **During assessment** β€” shown before the first question of each new domain - **After a wrong answer** β€” "Show me the handout" is one of the follow-up options - **During capstone build** β€” relevant handouts are available at each build step ### Generating PDF Handouts Generate branded PDFs for all 30 handouts for offline study: ```bash cd connectrylab-architect-cert-mcp npm run generate:pdfs ``` This creates 30 PDFs in `generated/handouts/` with: - Architect Cert logo and domain label in the header - Clean formatting with syntax-highlighted code blocks - "Connectry LABS β€” Claude Certified Architect Exam Prep β€” Free & Open Source" footer
## Adaptive Practice Every practice question is selected by a **three-priority algorithm**: 1. **Overdue reviews** β€” Spaced repetition items due for review today 2. **Weak areas** β€” Topics where your mastery is below 50% 3. **New material** β€” Fresh questions from your recommended domain ### Filtering You can filter practice questions by domain and/or difficulty: ``` You: "Give me a hard question about prompt engineering" Claude: [presents a hard Domain 4 question with clickable A/B/C/D buttons] ``` ### Interactive Follow-Ups After every answer, you get clickable follow-up options. Each option dives deeper into the concept β€” then brings you right back to your quiz. You never leave your study flow to look something up. ### Mastery Levels Each of the 30 task statements has an independent mastery level: | Level | Criteria | What it means | |-------|----------|---------------| | **Unassessed** | No attempts yet | You haven't seen questions on this topic | | **Weak** | < 50% accuracy | Needs significant study β€” questions resurface frequently | | **Developing** | 50-69% accuracy | Making progress β€” keep practicing | | **Strong** | 70-89% accuracy | Good understanding β€” review intervals are longer | | **Mastered** | >= 90% accuracy, 5+ attempts, 3+ consecutive correct | Exam-ready β€” rare reviews | ### Spaced Repetition (SM-2) The [SM-2 algorithm](https://en.wikipedia.org/wiki/SuperMemo#Description_of_SM-2_algorithm) schedules review intervals: - **First review:** 1 day after answering - **Second review:** 3 days after first review - **Subsequent reviews:** Previous interval x ease factor (starts at 2.5) - **Wrong answer:** Interval resets, ease factor decreases by 0.2 (floor: 1.3) - **Correct answer:** Ease factor increases by 0.1 Difficult questions come back often. Easy ones space out to weeks or months.
## Guided Capstone Build The most hands-on way to learn β€” 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. ### How It Works The capstone build has three phases: **Phase 1 β€” Project Shaping** You 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. ``` You: "I want to start a guided capstone build" Claude: [presents the 30 criteria across all 5 domains] Describe your project idea and I'll analyze coverage. You: "A multi-agent code review system that analyzes PRs" Claude: Your idea naturally covers 24/30 criteria. To cover the remaining 6, I'd suggest adding: [specific suggestions mapped to task statements] ``` **Phase 2 β€” Interleaved Build (18 steps)** Each step follows the same pattern: 1. **Quiz** β€” 2-3 questions on the task statements you're about to build (clickable A/B/C/D) 2. **Build** β€” Claude generates the file's code, themed to your project 3. **Walkthrough** β€” Line-by-line explanation mapping code to task statements A **visual 18-step checklist** tracks your progress in real time. The 18 steps build incrementally: | Steps | What you build | Task Statements | |-------|---------------|-----------------| | 1-2 | Project config (CLAUDE.md, package.json) | 3.1-3.4 | | 3-5 | MCP server, tools, error handling | 2.1-2.5 | | 6-10 | Agentic loop, subagents, hooks, workflows, sessions | 1.1-1.7 | | 11-13 | Prompts: system, extraction, batch processing | 4.1-4.6 | | 14-18 | Context: preservation, triggers, propagation, scratchpad, confidence | 5.1-5.6 | Every quiz answer feeds into the same spaced repetition and mastery tracking as regular practice. **Phase 3 β€” Final Review** After 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. ### Capstone Build Tools | Tool | What it does | |------|-------------| | `start_capstone_build` | See the 30 criteria, describe your theme, refine until coverage is complete | | `capstone_build_step` | Drive the build: confirm, quiz, build, next, status, or abandon | | `capstone_build_status` | Check your progress β€” current step, criteria coverage, quiz performance | ### How It Connects to Everything Else - Quiz answers during the build use the same `submit_answer` grading and SM-2 scheduling - After any quiz question, you can use the same follow-up options (code example, concept lesson, handout, reference project) - The reference projects show how the capstone structure looks when complete - Progress persists across sessions β€” pick up where you left off
## Practice Exams Full 60-question exams that simulate the real certification: | Detail | Value | |--------|-------| | Total questions | 60 | | D1: Agentic Architecture | 16 questions (27%) | | D2: Tool Design & MCP | 11 questions (18%) | | D3: Claude Code Config | 12 questions (20%) | | D4: Prompt Engineering | 12 questions (20%) | | D5: Context & Reliability | 9 questions (15%) | | Scoring | 0-1000, passing at 720 | | Question selection | Fresh set each time β€” avoids repeating your most recent attempt | | UI | Clickable A/B/C/D buttons with code previews | | Progress | Visual 60-item checklist updated after each answer | All attempts are saved with per-domain score breakdowns and improvement trends.
## Progress Dashboard Architect Cert includes a **glassmorphism visual dashboard** that renders directly inside Claude via Claude Preview. ### What It Shows - **Readiness Ring** β€” Overall exam readiness percentage in a circular progress indicator - **Domain Mastery Grid** β€” 5 cards with progress bars for each exam domain - **Exam History Chart** β€” Line chart plotting your practice exam scores over time, with the 720 passing score marked - **Recent Activity Timeline** β€” Your last 10 answers with correct/incorrect indicators - **Capstone Progress** β€” Current build step and completion percentage - **Quick Action Buttons** β€” Jump to practice, exam, study plan, or capstone ### How to Open ``` You: "Show my dashboard" Claude: [opens the branded dashboard in Claude Preview] ``` The dashboard can be reopened at any time by asking Claude again. It always reflects your latest data. ### Text Fallback For MCP clients that don't support Claude Preview, the dashboard tool also returns a **text summary** with mastery percentages and exam stats.
## Reference Projects Architect Cert includes **6 complete reference projects** β€” runnable TypeScript codebases that demonstrate certification concepts in real code. Every file has a header comment mapping it to specific task statements. | Project | Focus | Files | What You'll See | |---------|-------|-------|-----------------| | **Capstone** | All 5 domains | 24 | Full multi-agent support system with MCP server, coordinator, subagents, prompt engineering, context management, and hooks | | **D1 β€” Agentic Loop** | Domain 1 | 10 | Multi-agent research coordinator with agentic loops, subagent spawning, hooks, session management, and task decomposition | | **D2 β€” Tool Design** | Domain 2 | 12 | MCP server with split tools, structured errors, agent-scoped tool distribution, resources, and built-in tool patterns | | **D3 β€” Claude Code Config** | Domain 3 | 14 | Complete config reference: CLAUDE.md hierarchy, slash commands, path rules, CI/CD workflows β€” not runnable code, but a real config layout | | **D4 β€” Prompt Engineering** | Domain 4 | 11 | Data extraction pipeline with explicit criteria, few-shot, structured output, validation-retry, batch processing, and multi-pass review | | **D5 β€” Context Manager** | Domain 5 | 14 | Long-session patterns: context preservation, scratchpad, subagent delegation, escalation, error propagation, confidence calibration, provenance | ### How to Access ``` You: "Show me a reference project for domain 1" Claude: [returns the project README, file listing, and architecture walkthrough] ``` ### How They Connect to the Study Flow When you get a question wrong, one of the follow-up options is **"Show me in the reference project"** β€” 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.
## Study Plan Get personalized study recommendations based on your performance, exam weights, and spaced repetition schedule. ### Multi-Select Domain Focus When requesting a study plan, you can pick specific domains to focus on using **checkboxes**: ``` You: "What should I study next?" Claude: [shows domain checkboxes β€” select the ones you want to focus on] [generates a personalized study plan with a visual checklist] ``` The study plan creates a **visual checklist** so you can track your progress through each recommendation.
## Exam Domains The Claude Certified Architect β€” Foundations exam covers 5 domains: | # | Domain | Weight | Tasks | Questions | |---|--------|--------|-------|-----------| | 1 | Agentic Architecture & Orchestration | 27% | 7 | 91 | | 2 | Tool Design & MCP Integration | 18% | 5 | 65 | | 3 | Claude Code Configuration & Workflows | 20% | 6 | 78 | | 4 | Prompt Engineering & Structured Output | 20% | 6 | 78 | | 5 | Context Management & Reliability | 15% | 6 | 78 | | | **Total** | **100%** | **30** | **390** | ### 30 Task Statements
Domain 1 β€” Agentic Architecture & Orchestration (7 tasks, 91 questions) | Task | Description | |------|-------------| | 1.1 | Design and implement agentic loops for autonomous task execution | | 1.2 | Orchestrate multi-agent systems with coordinator-subagent patterns | | 1.3 | Configure subagent invocation, context passing, and spawning | | 1.4 | Implement multi-step workflows with enforcement and handoff patterns | | 1.5 | Apply Agent SDK hooks for tool call interception and data normalization | | 1.6 | Design task decomposition strategies for complex workflows | | 1.7 | Manage session state, resumption, and forking |
Domain 2 β€” Tool Design & MCP Integration (5 tasks, 65 questions) | Task | Description | |------|-------------| | 2.1 | Design effective tool interfaces with clear descriptions and boundaries | | 2.2 | Implement structured error responses for MCP tools | | 2.3 | Distribute tools appropriately across agents and configure tool choice | | 2.4 | Integrate MCP servers into Claude Code and agent workflows | | 2.5 | Select and apply built-in tools effectively |
Domain 3 β€” Claude Code Configuration & Workflows (6 tasks, 78 questions) | Task | Description | |------|-------------| | 3.1 | Configure CLAUDE.md files with appropriate hierarchy and scoping | | 3.2 | Create and configure custom slash commands and skills | | 3.3 | Apply path-specific rules for conditional convention loading | | 3.4 | Determine when to use plan mode vs direct execution | | 3.5 | Apply iterative refinement techniques for progressive improvement | | 3.6 | Integrate Claude Code into CI/CD pipelines |
Domain 4 β€” Prompt Engineering & Structured Output (6 tasks, 78 questions) | Task | Description | |------|-------------| | 4.1 | Design prompts with explicit criteria to improve precision | | 4.2 | Apply few-shot prompting to improve output consistency | | 4.3 | Enforce structured output using tool use and JSON schemas | | 4.4 | Implement validation, retry, and feedback loops | | 4.5 | Design efficient batch processing strategies | | 4.6 | Design multi-instance and multi-pass review architectures |
Domain 5 β€” Context Management & Reliability (6 tasks, 78 questions) | Task | Description | |------|-------------| | 5.1 | Manage conversation context to preserve critical information | | 5.2 | Design effective escalation and ambiguity resolution patterns | | 5.3 | Implement error propagation strategies across multi-agent systems | | 5.4 | Manage context effectively in large codebase exploration | | 5.5 | Design human review workflows and confidence calibration | | 5.6 | Preserve information provenance and handle uncertainty in synthesis |

## Tools Architect Cert provides **18 MCP tools** that Claude uses to deliver the study experience: ### Study Flow | Tool | Description | |------|-------------| | `start_assessment` | Begin with 15 diagnostic questions (lessons-first, one at a time) to determine your learning path | | `get_practice_question` | Get the next adaptive question with clickable A/B/C/D buttons (reviews > weak areas > new material) | | `submit_answer` | Grade your answer deterministically β€” presents interactive follow-up options | | `follow_up` | Handle post-answer actions: code examples, concept lessons, handouts, reference projects | | `get_section_details` | Deep dive into a specific task statement with full concept handout | ### Progress & Planning | Tool | Description | |------|-------------| | `get_progress` | View overall study progress with mastery percentages per domain | | `get_curriculum` | Browse all 5 domains and 30 task statements with current mastery levels | | `get_weak_areas` | Identify topics that need the most work, ranked by weakness | | `get_study_plan` | Get personalized recommendations with multi-select domain focus | | `get_dashboard` | Open the visual progress dashboard in Claude Preview | ### Practice Exams | Tool | Description | |------|-------------| | `start_practice_exam` | Take a full 60-question practice exam simulating the real certification | | `submit_exam_answer` | Submit and grade answers during a practice exam | | `get_exam_history` | View all past exam attempts with scores, trends, and per-domain comparison | ### Capstone Build | Tool | Description | |------|-------------| | `start_capstone_build` | Start a guided capstone build β€” shape your project and validate criteria coverage | | `capstone_build_step` | Drive the capstone build: confirm, quiz, build, next, status, or abandon | | `capstone_build_status` | Check capstone build progress β€” current step, coverage, quiz performance | ### Reference & Admin | Tool | Description | |------|-------------| | `scaffold_project` | Access reference projects for hands-on practice with real code | | `reset_progress` | Start over β€” requires explicit confirmation to prevent accidents | The server also registers **8 interactive prompts** and **3 resource types** (concept handouts, reference projects, exam overview).
## Architecture ``` β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β” β”‚ YOUR MACHINE β”‚ β”‚ β”‚ β”‚ β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β” β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β” β”‚ β”‚ β”‚ Claude Desktopβ”‚ β”‚ Architect Cert MCP β”‚ β”‚ β”‚ β”‚ Claude Code │◄─────►│ β”‚ β”‚ β”‚ β”‚ Any MCP clientβ”‚ stdio β”‚ 18 tools β”‚ β”‚ β”‚ β””β”€β”€β”€β”€β”€β”€β”¬β”€β”€β”€β”€β”€β”€β”€β”˜ β”‚ 8 prompts β”‚ β”‚ β”‚ β”‚ β”‚ 3 resource types β”‚ β”‚ β”‚ β”‚ β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”¬β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜ β”‚ β”‚ β”‚ β”‚ β”‚ β”‚ β”‚ β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”Όβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β” β”‚ β”‚ β”‚ β”‚ β”‚ β”‚ β”‚ β”‚ β”‚ ~/.connectry- 390 questions 6 reference β”‚ β”‚ β”‚ architect/ 30 handouts projects β”‚ β”‚ β”‚ progress.db (bundled JSON/MD) (bundled TS) β”‚ β”‚ β”‚ β”‚ β”‚ β”‚ β”Œβ”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β” β”‚ β”‚ └──► Claude Preview (Dashboard) β”‚ β”‚ β”‚ β”‚ localhost HTTP server β”‚ β”‚ β”‚ β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜ β”‚ β”‚ β”‚ β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜ ``` ### Core Components | Component | Technology | Purpose | |-----------|-----------|---------| | MCP Server | `@modelcontextprotocol/sdk` v1 | Registers tools, prompts, resources over stdio | | Grading Engine | Pure TypeScript functions | Deterministic answer verification | | Spaced Repetition | SM-2 algorithm | Optimal review scheduling | | Question Selector | Three-priority algorithm | Overdue reviews > weak areas > new material | | Follow-Up System | State-driven tool chain | Interactive post-answer detours | | Capstone Build Engine | 18-step interleaved builder | Guided learn-build-explain flow with LLM validation | | Dashboard Server | Node.js HTTP + HTML | Glassmorphism visual dashboard via Claude Preview | | Question Bank | 390 bundled JSON questions | Scenario-based, verified against docs | | Concept Handouts | 30 bundled markdown files | Structured study materials per task statement | | Reference Projects | 6 bundled TypeScript projects | Runnable code demonstrating each domain | | PDF Generator | Puppeteer + Marked | Branded handout PDFs for offline study | | Progress Store | `better-sqlite3` (WAL mode) | Persistent mastery, answers, schedules | ### Interactive UI Architecture Architect Cert doesn't build its own chat UI. Instead, it instructs Claude to use **built-in interactive tools**: - **AskUserQuestion** β€” Presents clickable buttons for A/B/C/D answers, follow-up actions, and domain selection. Supports single-select (radio buttons), multi-select (checkboxes), and free-text input via "Other". - **TodoWrite** β€” Creates visual progress checklists that update in real time during assessments, exams, and capstone builds. - **Claude Preview** β€” Renders the glassmorphism dashboard HTML on a local HTTP server. This approach works across Claude Code and Claude Desktop without requiring any custom UI code on the client side. ### Anti-Sycophancy Design This server enforces honest grading at the protocol level β€” not just in prompts: 1. **Deterministic grading** β€” `gradeAnswer()` is a pure function. No LLM is involved in judging correctness. 2. **Tool-level enforcement** β€” The `submit_answer` tool description instructs Claude to relay results verbatim. 3. **No partial credit** β€” Multiple choice, one correct answer. No "you were on the right track." 4. **Wrong answer explanations** β€” Every incorrect option has a specific `whyWrongMap` entry explaining the misconception. 5. **System prompt rules** β€” Five anti-sycophancy directives prevent Claude from softening incorrect results.
## Question Bank Details | Metric | Value | |--------|-------| | Total questions | 390 | | Domains covered | 5 | | Task statements covered | 30 | | Questions per task statement | 13 | | Difficulty distribution | ~4 easy, 5 medium, ~4 hard per task | | Answer key balance | Distributed across A/B/C/D | | Question format | Scenario-based multiple choice | | Each question includes | Scenario, question, 4 options, explanation, why-wrong-map, references | | Source material | Anthropic official documentation |
## Data Storage - Progress is stored locally at `~/.connectry-architect/progress.db` (SQLite, WAL mode) - Your user config lives at `~/.connectry-architect/config.json` (auto-created on first run) - No cloud, no accounts, no telemetry β€” everything stays on your machine
## Contributing We welcome contributions! Here's how to get started: ```bash # Clone the repo git clone https://github.com/Connectry-io/connectrylab-architect-cert-mcp.git cd connectrylab-architect-cert-mcp # Install dependencies npm install # Build npm run build # Run tests npm test # Generate PDF handouts npm run generate:pdfs # Run locally node dist/index.js ``` ### Project Structure ``` src/ β”œβ”€β”€ index.ts # MCP server entry point β”œβ”€β”€ config.ts # User config management β”œβ”€β”€ types.ts # All TypeScript interfaces β”œβ”€β”€ data/ β”‚ β”œβ”€β”€ loader.ts # Lazy-cached data loading β”‚ β”œβ”€β”€ curriculum.json # 30 task statements β”‚ β”œβ”€β”€ questions/ # 390 questions (5 domain files) β”‚ β”œβ”€β”€ handouts/ # 30 concept handouts (markdown) β”‚ β”œβ”€β”€ criteria.ts # 30 task statement criteria for capstone validation β”‚ β”œβ”€β”€ build-steps.ts # 18 capstone build step definitions β”‚ └── system-prompt.ts # Anti-sycophancy rules β”œβ”€β”€ db/ β”‚ β”œβ”€β”€ schema.ts # SQLite schema (9 tables) β”‚ β”œβ”€β”€ store.ts # Database initialization β”‚ β”œβ”€β”€ mastery.ts # Mastery level calculations β”‚ β”œβ”€β”€ answers.ts # Answer recording β”‚ β”œβ”€β”€ review-schedule.ts # SM-2 review scheduling β”‚ β”œβ”€β”€ capstone.ts # Capstone build CRUD operations β”‚ β”œβ”€β”€ users.ts # User management β”‚ └── exam-attempts.ts # Practice exam tracking β”œβ”€β”€ engine/ β”‚ β”œβ”€β”€ grading.ts # Deterministic grading β”‚ β”œβ”€β”€ spaced-repetition.ts # SM-2 algorithm β”‚ β”œβ”€β”€ question-selector.ts # Priority-based selection β”‚ β”œβ”€β”€ exam-builder.ts # Practice exam generation β”‚ └── adaptive-path.ts # Learning path recommendations β”œβ”€β”€ tools/ # 18 MCP tool handlers β”‚ β”œβ”€β”€ index.ts # Tool registration β”‚ β”œβ”€β”€ elicit.ts # MCP elicitation helper (graceful fallback) β”‚ β”œβ”€β”€ start-assessment.ts β”‚ β”œβ”€β”€ submit-answer.ts β”‚ β”œβ”€β”€ get-practice-question.ts β”‚ β”œβ”€β”€ follow-up.ts β”‚ β”œβ”€β”€ get-progress.ts β”‚ β”œβ”€β”€ get-curriculum.ts β”‚ β”œβ”€β”€ get-section-details.ts β”‚ β”œβ”€β”€ get-weak-areas.ts β”‚ β”œβ”€β”€ get-study-plan.ts β”‚ β”œβ”€β”€ start-practice-exam.ts β”‚ β”œβ”€β”€ submit-exam-answer.ts β”‚ β”œβ”€β”€ get-exam-history.ts β”‚ β”œβ”€β”€ scaffold-project.ts β”‚ β”œβ”€β”€ start-capstone-build.ts β”‚ β”œβ”€β”€ capstone-build-step.ts β”‚ β”œβ”€β”€ capstone-build-status.ts β”‚ β”œβ”€β”€ reset-progress.ts β”‚ └── dashboard.ts β”œβ”€β”€ ui/ β”‚ β”œβ”€β”€ server.ts # Dashboard HTTP server β”‚ β”œβ”€β”€ dashboard.html # Glassmorphism dashboard UI β”‚ β”œβ”€β”€ meta.ts # Quiz widget metadata β”‚ └── loader.ts # HTML asset loader β”œβ”€β”€ prompts/ # 8 MCP prompt definitions └── resources/ # 3 MCP resource types projects/ β”œβ”€β”€ capstone/ # All 5 domains β€” multi-agent support system β”œβ”€β”€ d1-agentic/ # Domain 1 β€” agentic loop research coordinator β”œβ”€β”€ d2-tools/ # Domain 2 β€” MCP server with tool patterns β”œβ”€β”€ d3-config/ # Domain 3 β€” Claude Code configuration layout β”œβ”€β”€ d4-prompts/ # Domain 4 β€” extraction & prompt engineering └── d5-context/ # Domain 5 β€” context management & reliability scripts/ └── generate-pdfs.ts # PDF handout generator ```
## License MIT Β© [Connectry Labs](https://connectry.io/labs)
## Credits - [Anthropic](https://anthropic.com) β€” Claude & the Claude Certified Architect certification program - [Model Context Protocol](https://modelcontextprotocol.io/) β€” The protocol that makes this possible - [better-sqlite3](https://github.com/WiseLibs/better-sqlite3) β€” Fast, synchronous SQLite for Node.js
---

Built with care by Connectry Labs

GitHub β€’ npm β€’ Architect Cert β€’ Connectry Labs