Model Context Protocol servers that extend AI capabilities
fr0ster
SAP AI Core LLM provider (ILlm) for @mcp-abap-adt/llm-agent.
fr0ster
OpenAI embedding provider (IEmbedderBatch) for @mcp-abap-adt/llm-agent.
fr0ster
Ollama embedding provider and OllamaRag convenience class for @mcp-abap-adt/llm-agent.
fr0ster
SAP AI Core embedding provider (IEmbedder) for @mcp-abap-adt/llm-agent.
fr0ster
HanaVectorRag vector store (SAP HANA Cloud Vector Engine) and HanaVectorRagProvider for @mcp-abap-adt/llm-agent.
fr0ster
PgVectorRag (PostgreSQL + pgvector) and PgVectorRagProvider for @mcp-abap-adt/llm-agent.
fr0ster
The SmartServer composition runtime as an **importable library**. It sits between the binary `@mcp-abap-adt/llm-agent-server` and the core `@mcp-abap-adt/llm-agent-libs`, so the SmartServer/coordinator composition can be reused in other projects without d
...morefr0ster
Default SmartAgent implementation with LLM providers, MCP client, HTTP server, and CLI.
xnuapp
Model Context Protocol server for Remoboard — let an AI assistant type into your phone.
xoai
Iris MCP-server channel (MCP dual-use) — the agent exposed AS an MCP server over JSON-RPC 2.0 (stdio). Same two-identifier protocol as channel-rest: the channel mints the sessionId, issues + rotates the continuationToken (atomically single-use), and surfa
...morealanripple
Local single-user MCP server for a TrainHeroic coach. No database, no Cloudflare deps.
Procyon Creative Services
Multi-account wrapper for Google Drive MCP with setup wizard and Claude-assisted automation
agentiknet
@agentproto/mcp-server — exposes every AIP doctype as MCP tools. Pass DoctypeSpec instances + an optional workspace dir for EXTENSION.md auto-loading; the server registers create_X / load_X / list_X / update_X / resolve_X / delete_X for each registered do
...morecodeform_ai
Codeform plugin to use your Claude Max subscription via Meridian proxy
xoai
Iris model_call performer — direct Anthropic Messages adapter via built-in fetch (zero deps). Real adapter runs only with ANTHROPIC_API_KEY; tests use a fake fetch / fake model.
...morexoai
Vendor-neutral, replay-safe model adapter — the OpenAI model_call performer behind Iris's portable model port; a peer of @irisrun/provider-anthropic behind the same port, so swapping providers leaves the agent and its byte-identical journal replay unchang
...moreReasonRDN — the shared memory substrate for the agentic era. Local-first with seamless auto-deposit to the warf Xchange (https://warf.astrognosy.com). reason:// handoffs, protected structural fingerprints (PCF), unified client, embedded node, first-class MCP, and CLI.
...moreaananda-giri
Whisper to free web LLM chat UIs (Qwen, ChatGPT, Claude, DeepSeek, GLM) and expose them over one local HTTP API — no API keys needed.
...moreCaucus contributors
Caucus MCP server over the backbone interface.
Mustafa Oransel
MCP server for NinjaTrader NinjaScript documentation - query the complete SDK reference
Paya Ebrahimi
Model Context Protocol server for the Pennivo markdown workspace
abluva-research
Composable data security plane for Model Context Protocol. Pluggable layers — collect, analyze, guard — across 50+ enterprise providers. Open architecture.
...moreweellio
A live, animated mission-control dashboard for your Claude Code agents — watch, manage, search, and control every agent (and the bill) from one screen.
...moreStache Team
Utilities and tools for the Stache RAG system