Intuition Spotlights: MCP

The Intuition Grants Program supports builders exploring new ways to create, curate, and leverage knowledge on top of the Intuition Protocol.

Meet Elijah Esin

Elijah Esin, known online as Ludarep, is a full-stack blockchain developer focused on trust infrastructure for AI agents and onchain systems.

His work sits at the intersection of reputation, graph intelligence, and autonomous systems. During Cohort 1, Elijah built the first working MVP of Intuition MCP and expanded it into a trust-scoring platform capable of evaluating relationships across large-scale attestation graphs.

His research and implementation work spans EigenTrust, AgentRank, graph databases, and practical trust systems designed for both humans and machines.

Outside of software, Elijah also operates a traditional business, giving him a perspective grounded in both technical experimentation and real-world execution.

Everyone wants to know who to trust onchain. I'm just trying to make that an actual question you can ask and get an answer to.

— Elijah Esin

Intuition MCP

Intuition MCP is a Model Context Protocol server that allows AI agents to query onchain trust and reputation using natural language.

Rather than forcing developers to build custom graph infrastructure, Intuition MCP provides a ready-made trust layer powered by the Intuition Knowledge Graph.

Using attestation data as its foundation, the system generates weighted reputation scores that can be consumed by agents, applications, and autonomous workflows.

The Problem

Most AI agents operate with little understanding of trust.

They can execute transactions, call APIs, and coordinate actions, but they often lack the ability to evaluate whether a wallet, user, or counterparty is trustworthy.

Without reputation awareness, agents are forced to operate blindly.

As autonomous systems become more common, the need for machine-readable trust infrastructure becomes increasingly important.

The Solution

Intuition MCP transforms raw attestation data into actionable trust signals.

The system ingests graph data into Neo4j and applies reputation algorithms such as EigenTrust and AgentRank to generate weighted, sybil-resistant trust scores.

Key capabilities include:

Composite reputation scores
Multi-hop trust analysis
Trust decay models
Reputation ranking
Natural language querying
Live testing playgrounds

The result is a trust engine that allows developers to ask meaningful questions about reputation without building the entire infrastructure stack themselves.

Building with Intuition

Elijah was drawn to Intuition because it approached a problem many teams discuss but few actually solve.

While onchain reputation has existed as a concept for years, Intuition provides the primitives necessary to make trust queryable, composable, and usable across applications.

For Intuition MCP, the protocol became the foundation for transforming reputation from an abstract idea into a practical developer tool.

Lessons Learned

One of the most valuable lessons came from customer discovery rather than engineering.

Through accelerator sessions and mentorship, Elijah learned the importance of validating usefulness before building complexity.

Shipping features is easy.

Understanding whether those features solve a meaningful problem is considerably harder.

The experience reinforced the importance of talking to users early and often.

What's Next?

As AI agents become more autonomous, Intuition MCP aims to become a core trust layer for the next generation of agent applications.

Future work includes expanding reputation methodologies, improving developer tooling, and making trust queries accessible across a wider range of agent frameworks.