Mkopo MCP Server

💳 mkopo-mcp — Alternative Credit Scoring MCP Server

Install: pip install mkopo-mcp · Use with any MCP client.

Most East Africans transacting heavily on mobile money have no formal credit file — the data exists in M-PESA history, but no scoring model reads it as an agent-callable tool.

70%+ of Kenyan adults are "credit invisible" — no formal credit history means no access to capital. M-PESA is a de facto financial history record that predicts repayment behaviour. mkopo-mcp makes that intelligence accessible to AI agents.

The Structural Problem

In mature economies, credit bureaus have 30+ years of credit card, mortgage, and auto loan data per person. That data machine took 50 years to build. In Kenya, most people's financial lives happen in cash and mobile money — invisible to formal credit systems.

M-PESA is the credit bureau that doesn't know it is one.

• Regular income deposits → income stability
• Utility payment regularity → discipline signal
• Savings behaviour (M-Shwari, chama) → future orientation
• Fuliza usage ratio → short-term distress indicator
• Business Paybill receipts → verifiable commercial activity

Tools

Quick Start

pip install mkopo-mcp       # coming soon to PyPI
mkopo-mcp                   # starts on stdio

Research Basis

• World Bank "Expanding Access to Finance" (2022)
• Breza & Kinnan "Measuring the Equilibrium Impacts of Credit" (2021)
• CBK Prudential Guidelines (2022)
• Kenya CRB Act (Cap. 44A)

⚠️ DEMO data — not a real CRB product. Consult Metropol or CRB Africa for official reports.

---

© 2026 Gabriel Mahia / AI Kung Fu LLC · MIT License

Live data

• kenya_macro_indicators: Kenya's annual inflation (IMF, projections flagged) and lending and deposit rates (World Bank, CC BY 4.0).

Each live answer names its source, attribution and retrieval time; if a service does not answer, the tool returns an error instead of a number. The other tools are unchanged.

Part of the East Africa Coordination Stack

This MCP server is part of the Kenya coordination infrastructure. Connect it to africa-coord-bus — the coordination event bus that routes signals between domains automatically.

pip install africa-coord-bus

All servers: pypi.org/user/gmahia Live demo: coord-cascade-demo

IP & Collaboration

MIT licensed. Feedback via GitHub Issues only — pull requests are not accepted. Demo data is labeled DEMO and is not suitable for operational decisions. Full policy: docs/architecture/IP_POLICY.md. Security reports: see SECURITY.md.

Part of the East Africa coordination stack

• Install & run: pip install reli-cli && reli list — the MCP servers on the official MCP Registry under io.github.gabrielmahia
• Evaluate any model on Swahili agent tasks: kipimo · dataset · leaderboard
• Coordinate across servers: africa-coord-bus — offline-first event bus with a built-in Kenya routing table
• Datasets: huggingface.co/gmahia · Docs hub: nairobi-stack

Model-agnostic by design: closed APIs, open-weight models, and small distilled models are all first-class citizens.

Demonstration only, and Kenya's rules on automated decisions

These tools produce synthetic demonstration results, not credit decisions, and they are not a Credit Reference Bureau product. Do not pass a real person's name or ID number to them.

If you build something real on a score like this, note the law: in Kenya the Data Protection Act 2019 is in force, including its provisions on decisions made solely by automated processing (section 35). The Artificial Intelligence Bill 2026 is a Senate bill and is not law: a Senate committee recommended abandoning it in October 2026, so as of 2026-10-23 Kenya has no dedicated AI law and the Data Protection Act is what applies. The bill would have added a right to an explanation and to human review of significant automated decisions such as a loan rejection, and commentators expected credit scoring to be treated as high-risk under such a law. Every scoring answer from this server carries an automated_decision_notice saying so. The lender examples are generic lender types, not named companies: an earlier version listed real lenders by tier, which this synthetic score cannot support.