[PR #533] [MERGED] Add Multi-MCP Agent Forge - specialized agents with MCP tool routing #2463

New Issue

GiteaMirror · 2026-04-21T08:32:30-05:00

GiteaMirror commented

2026-04-21 08:32:30 -05:00

📋 Pull Request Information

Original PR: https://github.com/Shubhamsaboo/awesome-llm-apps/pull/533
Author: @WeberG619
Created: 2/25/2026
Status: ✅ Merged
Merged: 3/7/2026
Merged by: @Shubhamsaboo

Base: main ← Head: add-multi-mcp-agent-forge

📝 Commits (2)

59bfefa Add Multi-MCP Agent Forge - specialized agents with MCP tools
bbb2183 feat: wire up actual MCP server connections for agent tool use

📊 Changes

3 files changed (+455 additions, -0 deletions)

View changed files

➕ mcp_ai_agents/multi_mcp_agent_forge/README.md (+80 -0)
➕ mcp_ai_agents/multi_mcp_agent_forge/agent_forge.py (+371 -0)
➕ mcp_ai_agents/multi_mcp_agent_forge/requirements.txt (+4 -0)

📄 Description

What This Adds

A complete, self-contained Streamlit app (mcp_ai_agents/multi_mcp_agent_forge/) that demonstrates the multi-agent + MCP pattern:

4 specialized agents (Code Reviewer, Security Auditor, Researcher, BIM Engineer)
Each agent connects to different MCP servers based on its domain
Automatic query routing classifies user intent and selects the best agent
Uses the Anthropic SDK with Claude for inference

Files

File	Description
`agent_forge.py`	Main Streamlit app (~200 lines)
`requirements.txt`	Dependencies (streamlit, anthropic, mcp, pydantic)
`README.md`	Setup instructions, architecture diagram, extension guide

How to Run

cd mcp_ai_agents/multi_mcp_agent_forge
pip install -r requirements.txt
streamlit run agent_forge.py

Why This Pattern Matters

Instead of giving one agent 50+ tools (which degrades tool selection accuracy), this routes queries to specialized agents that each have access to only the MCP tools they need. This is the core pattern behind cadre-ai.

Previous Submission

This replaces #507 which was closed because it only linked to an external repo. This PR includes the full self-contained code.

_{🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.}

## 📋 Pull Request Information **Original PR:** https://github.com/Shubhamsaboo/awesome-llm-apps/pull/533 **Author:** [@WeberG619](https://github.com/WeberG619) **Created:** 2/25/2026 **Status:** ✅ Merged **Merged:** 3/7/2026 **Merged by:** [@Shubhamsaboo](https://github.com/Shubhamsaboo) **Base:** `main` ← **Head:** `add-multi-mcp-agent-forge` --- ### 📝 Commits (2) - [`59bfefa`](https://github.com/Shubhamsaboo/awesome-llm-apps/commit/59bfefabb126b4096323ee3714564a59f09991e2) Add Multi-MCP Agent Forge - specialized agents with MCP tools - [`bbb2183`](https://github.com/Shubhamsaboo/awesome-llm-apps/commit/bbb2183a119bc1be57e931624e88a5a5c792fc9b) feat: wire up actual MCP server connections for agent tool use ### 📊 Changes **3 files changed** (+455 additions, -0 deletions) <details> <summary>View changed files</summary> ➕ `mcp_ai_agents/multi_mcp_agent_forge/README.md` (+80 -0) ➕ `mcp_ai_agents/multi_mcp_agent_forge/agent_forge.py` (+371 -0) ➕ `mcp_ai_agents/multi_mcp_agent_forge/requirements.txt` (+4 -0) </details> ### 📄 Description ## What This Adds A complete, self-contained Streamlit app (`mcp_ai_agents/multi_mcp_agent_forge/`) that demonstrates the **multi-agent + MCP** pattern: - **4 specialized agents** (Code Reviewer, Security Auditor, Researcher, BIM Engineer) - Each agent connects to **different MCP servers** based on its domain - **Automatic query routing** classifies user intent and selects the best agent - Uses the **Anthropic SDK** with Claude for inference ## Files | File | Description | |------|-------------| | `agent_forge.py` | Main Streamlit app (~200 lines) | | `requirements.txt` | Dependencies (streamlit, anthropic, mcp, pydantic) | | `README.md` | Setup instructions, architecture diagram, extension guide | ## How to Run ```bash cd mcp_ai_agents/multi_mcp_agent_forge pip install -r requirements.txt streamlit run agent_forge.py ``` ## Why This Pattern Matters Instead of giving one agent 50+ tools (which degrades tool selection accuracy), this routes queries to specialized agents that each have access to only the MCP tools they need. This is the core pattern behind [cadre-ai](https://github.com/WeberG619/cadre-ai). ## Previous Submission This replaces #507 which was closed because it only linked to an external repo. This PR includes the full self-contained code. --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>

GiteaMirror added the pull-request label 2026-04-21 08:32:30 -05:00

GiteaMirror closed this issue

2026-04-21 08:32:30 -05:00

Sign in to join this conversation.

1 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: github-starred/awesome-llm-apps#2463