📡 MCP Radar

Andes-Software-Solutions/Andes.Extensions.AI

Composable Microsoft.Extensions.AI IChatClient middlewares for .NET — tool-call tracking with in-band streaming progress, per-tool token usage reports, and satellite packages for MCP tools and Microso

⭐ 5 stars
⑂ 2 forks
C#
momentum ▲ 14.0
created 2026-07-21
on radar since 2026-07-26
star trend 4 → 5 since 2026-07-26
csharpdotnetichatclientmicrosoft-agent-frameworkmicrosoft-extensions-aimodel-context-protocolnuget
View on GitHub ↗Homepage ↗

About Andes.Extensions.AI

Middleware extensions for Microsoft.Extensions.AI: per-request and per-tool token usage tracking, and streaming status propagation for IChatClient pipelines.

- Token usage tracking — input/output/total tokens for the main assistant, per model turn, and attributed to each tool call (including LLM calls nested inside tools), rolled up into a ChatUsageReport. - Streaming progress statuses — synthetic ChatProgressContent updates interleaved into the live stream so your UI can show "Calling GetWeather Tool", sub-statuses reported from inside the tool ("Extracting…", "Processing…"), and completion — while the model and tools are still working. - Out-of-band observers — implement IChatProgressObserver to receive the same events and the final report without parsing the stream. - Privacy by default — progress events never carry prompt content, tool arguments, or tool results unless explicitly opted in.

Register the middleware before UseFunctionInvocation() — the tracker must wrap the tools that the function-invoking client executes:

Tools that are themselves LLM-backed (for example, agents exposed as functions) can attribute their own usage to the calling scope with ChatProgress.ReportUsage(...) — or simply run their own UseToolTracking() pipeline, whose total rolls up automatically.

Before persisting responses into conversation history, remove the synthetic content:

First-class MCP support ships as a satellite package so the core stays dependency-lean:

From the project README.

Maintaining this server?

Add the radar badge to your README — it shows your project was picked up by MCP Radar and links to this page:

[![On MCP Radar](https://mcp.liqiwa.com/badge.svg)](https://mcp.liqiwa.com/s/Andes-Software-Solutions--Andes.Extensions.AI.html)

Related MCP servers

keithadler/leanstudio

A native desktop IDE for Lean 4 on macOS, Windows and Linux. Every proof re-checked by Tenet, an independent Lean 4 kernel. An MCP server for AI assistants, with Git and GitHub built in.

⭐ 5C#
peopleworks/XAFLogicExplainer

Teach your AI coding agent what your DevExpress XAF application actually does. Roslyn-based extraction of entities, controllers, business rules and Model Editor customizations.

⭐ 4C#
smollini/AgentQueueMcp

Firewall-friendly messaging for AI agents: MCP server that turns one Azure Storage account into an addressed multi-agent message bus. Pull-only, zero inbound ports.

⭐ 3C#
HorizunGroup/naviscoord-mcp

Open-source MCP server for Autodesk Navisworks Manage 2024-2026: AI-assisted BIM clash detection, root-cause analysis, coordination plans and verified model workflows.

⭐ 12C#
ashafizullah/rpt-mcp

MCP server for SAP Crystal Reports .rpt files: inspect and edit reports from AI assistants (Claude Code, Claude Desktop, Cursor)

⭐ 5C#
pinqponq/pinqops

Self-hosted deploy ops for closed servers — Docker, GitHub Actions, optional Caddy UI.

⭐ 4C#

📬 Get the weekly radar in your inbox

The top new MCP servers of the week, every Monday. No spam, unsubscribe anytime.