DocsIntroduction

Introduction

Opportunity infrastructure for AI agents.

An open-source MCP server for discovering, understanding, and acting on opportunities. ROUTE gives AI agents and developers structured access to opportunities across different categories and sources through a common interface.

Open source · Agent agnostic · MCP infrastructureRepository

Why ROUTE

The problem

Opportunities are everywhere. Access is fragmented.

Jobs live on job boards. Hackathons live on developer platforms. Scholarships, grants, fellowships, internships, and accelerators are spread across organization websites and specialized platforms.

This fragmentation makes it difficult for people and AI agents to discover opportunities, retrieve their details, and understand what it takes to pursue them.

An agent may understand a user's goals, but it still needs access to the opportunity itself. Every external platform introduces its own data formats, pages, and information structure.

ROUTE addresses this access problem with a common interface for opportunity discovery, retrieval, saving, and preparation. The goal is not another job board. It is infrastructure that other applications can build on.

How to use ROUTE

2 ways to build

One infrastructure layer. Two ways to build with it.

ROUTE connects opportunity sources to AI-powered experiences. You can use it through a compatible AI application or integrate its capabilities into an application of your own.

For usersAI assistant

Find opportunities with your AI assistant.

Connect a compatible MCP client to a running ROUTE server. Your assistant can then use ROUTE's tools to discover opportunities, retrieve their details, save interesting options, and gather preparation context.

01
Connect
Configure your AI application to connect to ROUTE.
02
Discover
Ask your assistant to find opportunities that match your request.
03
Explore and save
Retrieve details, compare options, and save opportunities for later.
04
Prepare
Use available opportunity context to understand requirements and plan your next steps.

Your AI assistant provides the conversation and reasoning. ROUTE provides the opportunity tools and data. Further automation depends on the capabilities and permissions of the connected agent.

Connect ROUTE

For developersStreamable HTTP

Build opportunity-powered applications.

Integrate ROUTE into your own AI application, agent, or workflow. Use its MCP tools as a focused opportunity layer while keeping your own application logic and user experience under your control.

01
Run ROUTE
Start the MCP server locally or deploy an accessible server.
02
Connect your client
Configure your MCP-compatible client to communicate through Streamable HTTP.
03
Compose capabilities
Use discovery, retrieval, saving, and preparation tools within your application or agent workflow.
04
Extend the system
Build your own experience or contribute providers and capabilities to the opportunity ecosystem.

ROUTE handles opportunity access. Your application remains responsible for its own model, business logic, user experience, authentication, and additional workflow orchestration.

Start building

Architecture overview

MCP · providers

One interface. Multiple opportunity sources.

ROUTE uses the Model Context Protocol (MCP) to expose tools to compatible AI applications. Its provider architecture handles source-specific retrieval and normalization behind that interface.

AI assistants

Conversational experiences

IDE agents

Developer workflows

Custom applications

Purpose-built experiences

MCP · Streamable HTTP

ROUTE MCP Server

Discover · Understand · Save · Prepare

Opportunity providers

External platforms and websites that publish opportunities across different categories.

Remote OKJob opportunity source

DevpostHackathon opportunity source

Source data is normalized into ROUTE's opportunity model.

Learn about the Model Context Protocol

Opportunity lifecycle

6 stages · 4 implemented

From discovery to taking action.

ROUTE is organized around the stages involved in pursuing an opportunity. Its core discovery and preparation capabilities are implemented, while broader action and tracking workflows remain part of the longer-term direction.

01
Discover
Implemented
Search supported opportunity sources and receive structured, normalized results that agents can work with.
Example: "Find remote jobs and upcoming hackathons."
02
Understand
Implemented
Retrieve an opportunity's details, including its description, organization, deadline, location, source, and other available information.
Example: "Tell me more about this opportunity."
03
Save
Implemented
Save opportunities for later use. Saved opportunity state is separate from the global sources used for discovery.
Example: "Save this opportunity for me."
04
Prepare
Implemented
Retrieve the main preparation context needed to get started with an opportunity. Planned sub-tools will extend this capability by retrieving deeper information from additional pages and resources.
Example: "What requirements should I understand before applying?"
05
Act
Future direction
Enable connected agents and applications to use opportunity context when helping users draft applications, plan projects, and work through next steps.
Example: "Help me prepare a strong application for this opportunity."
06
Track
Future direction
Explore workflows for tracking applications, deadlines, progress, opportunity updates, and follow-up actions.
Example: "Help me keep track of the opportunities I am pursuing."

Deeper opportunity context

The current prepare_opportunity tool retrieves the main preparation context needed to get started. Planned sub-tools will extend this capability by retrieving deeper information that may be spread across separate pages or resources. Depending on the opportunity, this could include eligibility criteria, application instructions, required documents, judging criteria, rules, or supporting resources. The connected agent can use that information to help the user make informed decisions and plan their next steps.

Opportunity coverage

9 categories · 2 in initial scope

One infrastructure layer. Many kinds of opportunities.

ROUTE is designed to support a broad range of opportunities through a consistent interface. The initial implementation focuses on jobs and hackathons, with an architecture designed to accommodate additional categories and providers over time.

JobsEmployment, remote work, and career opportunities.Initial scope

HackathonsCompetitions, challenges, and build events.Initial scope

ScholarshipsFunding opportunities for education.Planned

GrantsFunding for projects, research, and initiatives.Planned

FellowshipsAcademic, professional, and research programs.Planned

InternshipsEarly-career and practical work experience.Planned

AcceleratorsPrograms that help startups and founders grow.Planned

FreelanceContract work and independent projects.Planned

And moreAdditional categories as the provider ecosystem grows.Planned

Categories beyond jobs and hackathons represent the broader vision. Their availability will depend on the providers implemented.

Design principles

4 principles

A small core built to grow.

ROUTE is designed around a focused interface and an extensible architecture, so new sources and capabilities can be introduced without tying the system to one platform or AI application.

MCP first

ROUTE is fundamentally an MCP server. Its interfaces and integrations consume the infrastructure rather than define it.

Agent agnostic

ROUTE is designed to work with compatible MCP clients instead of requiring a particular AI assistant.

Provider independent

External platforms are isolated behind providers that normalize their information into a common opportunity model.

Composable capabilities

Focused tools can be combined by agents into larger workflows without turning ROUTE into a monolithic agent.

Getting started

3 steps

Connect your agent to ROUTE.

Start with the setup guide, connect a compatible MCP client, and explore the tools available in the current implementation.

01Quickstart

Set up ROUTE and learn how to make your first request.

02Connect an MCP client

Learn how a compatible AI application connects to ROUTE.

03Explore the tools

Understand the available tools, their inputs, and their outputs.

Continue to Architecture…