Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@StudyMate AI Build me a study plan for my exams next week"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
StudyMate helps students organize study tasks, plan exam preparation, understand workload risks, and recover from missed study sessions.
Built for the Amazon Developer Hackathon 2026 — Alexa+ Track, StudyMate exposes study operations through a self-hosted Model Context Protocol (MCP) server.
Current implementation: StudyMate uses deterministic application logic for planning, recovery, and deadline calculations. An LLM or Gemini-powered assistant is not currently integrated.
StudyMate focuses on three practical questions:
What needs to be done? Organize tasks, exams, and estimated effort.
What can realistically fit? Compare workload with available study time.
What happens when a plan breaks? Propose a recovery plan around unfinished work.
The architecture separates the user interface, application orchestration, study logic, MCP integration, and persistence. This keeps the study operations reusable without coupling them to one interface.
flowchart TB
subgraph USER["01 · USER EXPERIENCE"]
U["Student"]
UI["StudyMate Web Workspace<br/>Tasks · Plans · Exams · Recovery"]
U <--> UI
end
subgraph APP["02 · APPLICATION LAYER"]
API["Express Application<br/>HTTP Routes · Request Handling"]
VALID["Input Validation<br/>Structured Data"]
API <--> VALID
end
subgraph CORE["03 · STUDY OPERATIONS ENGINE"]
direction TB
TASK["Task & Exam Management"]
PLAN["Study Plan Proposals"]
RISK["Deadline Risk Calculations"]
REC["Session Recovery"]
APPROVE["Proposal Review<br/>Student Approval"]
TASK --> PLAN
PLAN --> RISK
RISK --> APPROVE
APPROVE -->|Approved plan| TASK
APPROVE -->|Recovery needed| REC
REC --> PLAN
end
subgraph MCP["04 · MCP INTEGRATION"]
CLIENT["MCP Client / Simulator"]
SERVER["MCP Server<br/>Streamable HTTP"]
TOOLS["Registered Tools<br/>list_tasks · add_task<br/>complete_task<br/>propose_study_plan<br/>list_exams"]
CLIENT <-->|MCP Requests & Responses| SERVER
SERVER <--> TOOLS
end
subgraph DATA["05 · PERSISTENCE"]
STORE[("Local JSON Storage<br/>data.json")]
end
UI <--> API
VALID <--> CORE
TOOLS <--> CORE
CORE <--> STORE
classDef user fill:#172554,stroke:#60A5FA,color:#FFFFFF,stroke-width:2px
classDef app fill:#164E63,stroke:#22D3EE,color:#FFFFFF,stroke-width:2px
classDef core fill:#312E81,stroke:#A5B4FC,color:#FFFFFF,stroke-width:2px
classDef mcp fill:#064E3B,stroke:#34D399,color:#FFFFFF,stroke-width:2px
classDef data fill:#713F12,stroke:#FBBF24,color:#FFFFFF,stroke-width:2px
class U,UI user
class API,VALID app
class TASK,PLAN,RISK,REC,APPROVE core
class CLIENT,SERVER,TOOLS mcp
class STORE dataRelated MCP server: StudyFlow AI MCP Server
Feature | Purpose |
Study Planning | Proposes plans based on task estimates and available time |
Deadline Risk Radar | Compares workload against available time |
Session Recovery | Replans around unfinished work |
Task & Exam Tracking | Organizes preparation |
Persistent State | Stores application data locally |
Student Approval | Keeps proposed changes under user control |
Endpoint: POST /mcp
MCP Tool | Purpose |
| Retrieve tasks |
| Add a task |
| Complete a task |
| Propose a study plan |
| Retrieve exams |
MCP lets compatible clients discover and invoke structured application tools. The project currently provides a self-hosted MCP server and web demonstration; it is not a deployed Alexa+ device integration or an Amazon-certified Agent Skill.
sequenceDiagram
autonumber
participant C as MCP Client / Simulator
participant S as StudyMate MCP Server
participant V as Validation
participant E as Study Operations Engine
participant D as Local Storage
C->>S: Initialize MCP session
S-->>C: Protocol capabilities
C->>S: Discover available tools
S-->>C: Tool definitions and schemas
C->>S: Call a supported tool
S->>V: Validate tool arguments
V->>E: Execute valid operation
E->>D: Read or update application state
D-->>E: Stored data or result
E-->>S: Operation result
S-->>C: MCP tool responseThe sequence shows the intended interaction pattern. Actual responses and available capabilities depend on the implemented server.
Node.js · Express.js · MCP SDK · Zod · dotenv · HTML/CSS/JavaScript · JSON storage
Requirements: Node.js 20+ and npm.
git clone https://github.com/shikharxcodes/amzan_hack.git
cd amzan_hack
npm install
npm startOpen http://localhost:3000 if using the default port.
Run tests:
npm testCriterion | Project evidence |
Technical Implementation | MCP tools and protocol interactions |
Design | Organized study workspace |
Potential Impact | Study planning and missed-session recovery |
Quality of Idea | A student-focused workflow exposed through MCP |
These are intended areas of alignment, not guaranteed judging scores.
No AI/LLM integration currently.
Local JSON storage; designed for single-user use.
No production authentication or hosted deployment.
Future work: optional AI study assistance, multi-user storage, and additional integrations.
The submission should include honest feedback on the tools, APIs, and SDKs actually used, including the MCP SDK, Node.js, Express.js, and validation/configuration libraries.
Criterion | StudyMate's Strength |
Technical Implementation | MCP server, tool discovery, structured tool calls |
Design | Organized, student-friendly study workspace |
Potential Impact | Practical study planning and session recovery |
Quality of Idea | Combines workload analysis, planning, and recovery |
Engineering highlights: Modular architecture · Explainable calculations · User-approved plans · Persistent state
MIT License · Copyright © 2026 Shikhar Maurya.
This server cannot be deployed
Streamable HTTP MCP server exposing planner flows, tasks, and squads.
Remote MCP server exposing SMI Aware tools, resources, and skills over Streamable HTTP.
Streamable HTTP MCP server for Google Calendar and Sheets with OAuth login.
Query, browse, and automate OmegaAI workspaces from any MCP client. Streamable HTTP with OAuth 2.0.
