Enterprise DNA
M MCP Servers Developer low

gabrielmahia/mpesa-mcp

by Various

MCP server for M-Pesa (Safaricom Daraja) and Africa's Talking — give AI agents East African fintech superpowers

G

MCP

gabrielmahia/mpesa-mcp

Added 11 June 2026

#ai-agents #daraja #fintech #kenya #mcp #mpesa #open-source #pypi

Overview

MCP server that wraps the M-Pesa (Safaricom Daraja) and Africa's Talking APIs into the Model Context Protocol. This allows AI agents to programmatically initiate payments, query balances, and send messages within East African fintech ecosystems.

Best for

Best for
Developers building AI agents that interact with East African mobile money and SMS services.

Use cases

  • Trigger M-Pesa payments from AI agent workflows
  • Query transaction history via Africa's Talking SMS
  • Automate mobile money operations for customer service bots

Notes

MCP server that wraps the M-Pesa (Safaricom Daraja) and Africa’s Talking APIs into the Model Context Protocol. This allows AI agents to programmatically initiate payments, query balances, and send messages within East African fintech ecosystems.

1 stars on GitHub. Last updated 2026-06-05. Licensed MIT.

Use cases

  • Trigger M-Pesa payments from AI agent workflows
  • Query transaction history via Africa’s Talking SMS
  • Automate mobile money operations for customer service bots

Pros

  • Direct MCP integration for standard AI agent access
  • Combines two popular East African fintech APIs in one server
  • Written in Python, easy to extend

Cons

  • Very early stage with minimal community adoption (1 star)
  • Limited documentation and may lack comprehensive error handling
  • Dependence on external API credentials and M-Pesa sandbox setup

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.

Pros

  • Direct MCP integration for standard AI agent access
  • Combines two popular East African fintech APIs in one server
  • Written in Python, easy to extend

Cons

  • Very early stage with minimal community adoption (1 star)
  • Limited documentation and may lack comprehensive error handling
  • Dependence on external API credentials and M-Pesa sandbox setup