"Connecting external data sources for contextual LLM calls" matching MCP servers:

Enables agents to detect and redact PII and secrets in text, HTML, and PDF documents before they reach an LLM, with reversible placeholder, mask, or hash redaction, and to list supported detectors and redaction styles.

Bridges Codex to external CLI coding agents for read-only code analysis and deterministic quality fixes.

Enables AI agents to query free public APIs — SEC EDGAR, World Bank, GDELT, DATA.GOV.HK, HKMA, the official MCP Registry, UK Companies House, FRED and Jina Reader — as zero-dependency MCP tools. Every result is wrapped in a { source, retrievedAt, data } provenance envelope so agents can cite the origin and time of each fact.

Provides RAG (Retrieval-Augmented Generation) capabilities via Contextual AI, enabling query processing and context-aware responses with citations. Integrates with MCP clients like Cursor IDE and Claude Desktop.

Routes work from your coding agent across 20+ free-tier LLMs to cut costs 10-100x using a round-robin dispatch pool with cooldown-aware auto-fallback.

Blind multi-model councils with anonymous LLM seats and direct 1:1 chat as a local MCP server. Enables asking one question to get independent answers from multiple model CLIs, then blind scoring and reveal.

Lets Claude Code send prompts to a locally running Qwen model on Apple Silicon and get answers back, with optional thinking mode for harder problems. It also lists the registered models and reports whether each one is running.

Enables an LLM to autonomously maintain a Markdown-based wiki with YAML frontmatter, using hybrid BM25 and local CPU vector search for fully offline retrieval without external APIs or databases.

Enables a primary agent to offload bounded, high-volume analysis of explicitly selected workspace files to an auxiliary LLM through a single read-only tool, returning text, Markdown, JSON, or unified-diff output with token usage while retaining all planning, verification, and edits.

Enables coding agents to query the current free LLM model lineup mid-task, check what free models were added or removed since the last scan, and look up curated provider free tiers with limits and caveats, so they stop hard-coding model IDs that silently break.

MCP server that provides a live coordination layer for AI agents, including attributable handoffs, a shared event ledger, atomic work-claiming, and advisory file leases to prevent collisions.

This MCP server enables MCP clients such as Claude Desktop to run LLM-generated code in a secure, isolated sandbox across multiple container backends and programming languages.

MCP server toolkit with payload library, WAF bypass generator, rate limiter, plugin system, and inter-MCP bridge for security testing.

Integrates Dify's external knowledge base API with the Model Context Protocol to enable AI agents to retrieve and query relevant information. It supports relevance scoring, metadata filtering, and flexible configuration through environment variables or command-line arguments.

A local-first memory cartridge for MCP-compatible AI clients, providing durable, inspectable, and auditable memory storage with a read-only-by-default MCP server.

Lets MCP clients delegate private or bulk text tasks to a local OpenAI-compatible LLM while keeping data on the machine.

MCP server that completes prompts with automatic provider fallback across OpenAI, Anthropic Claude, and Google Gemini. It is rate-limit aware, retries on transient failures, and provides a transparent attempt log.

Independent latency and uptime data for 46 hosted AI inference APIs, probed every 5 minutes from 4 regions with no gateway, plus a verified model deprecation calendar and Claude Code / Codex speed and accuracy drift. Four read-only tools, no API key; data CC BY 4.0.

MCP server that provides tools for evaluating LLM agent reliability, including adversarial task generation, automated LLM-as-judge assessment, and confidence statistics.

An MCP server that provides local, private, synchronous access to Ollama models for prompt answering, text classification, and model listing, without any file access or command execution.