"Using llama.cpp and Searxng with local models" matching MCP servers:

Free web search for your AI using public SearXNG instances, with parallel multi-server requests and throttling.

Student-focused Canvas LMS MCP server with 209 tools: what-if grades, the score you need on the final, deadlines and missing work across every course, submissions with confirm-before-send, discussions and inbox. Runs locally with npx, nothing exposed to the internet.

Enables local AI agents to generate and edit Figma designs via the Plugin API through a WebSocket relay, with no rate limits and all communication remaining local.

Enables fast literal and regex search over local workspaces using a sparse n-gram index for candidate pruning, with deterministic verification against file contents.

MCP server for AI-powered web research — search (Google, Brave, Serper, SearXNG), scrape any page, extract PDFs/DOCX/YouTube transcripts, academic & patent search. Single Go binary. Works with Claude, Cursor, Copilot, and any MCP client.

Enables AI agents to search local Markdown documents using natural language, with automatic indexing and section-level retrieval.

Enables writing to Figma from MCP clients like Claude Code without a Figma Dev seat, using a local WebSocket relay and a Figma dev plugin.

Enables AI assistants to read and modify Figma designs programmatically through the Model Context Protocol.

A lightweight MCP server that enables AI assistants to execute local development tools and retrieve system status with low latency over stdio or HTTP.

An MCP server that provides AI clients with access to developer documentation via llms.txt files. Exposes tools, resources, and prompts.

Deterministic encode, decode, hash and convert for agents, on the same pipeline engine as the String Utility Belt web app, CLI and editor extensions: a pipeline built and checked in the browser runs unchanged from its share link. Base64/hex/URL, JWT, hashes, JSON/YAML/CSV/XML, dates and regex; tools to search, describe and run utilities or whole pipelines, and to detect input formats. Runs locally

Enables AI assistants to perform Google searches with real, resolvable sources and transparent search terms.

MCP server for overseer.nvim — gives any MCP-capable coding agent running inside Neovim (Claude Code, Codex, Gemini CLI, opencode) the same visibility and control over tasks the user has: list, run, tail, restart, stop. Agent-started dev servers land in the user's own task list instead of becoming invisible orphaned processes.

Lets AI assistants discover public Socrata open-data portals, search their datasets, inspect schemas, and run bounded read-only SoQL queries to answer questions with live government data. All access is strictly read-only and results come straight from the publisher's portal.

Enables AI models to access and search offline Wikipedia and other knowledge bases stored in ZIM format files. Provides intelligent content retrieval, structured browsing, advanced search capabilities, and metadata extraction for comprehensive offline knowledge access.

Read-only MCP server for Grafana Loki: lets AI agents (Claude Code, Cursor, Copilot) discover labels, read bounded LogQL pages, count lines by label or time, and export logs — across several Loki stands

Enables multi-level reasoning with configurable depth levels, session-based state management, structured thought input, and real-time trace resources.

Relays error logs from other MCP servers when an LLM call fails and the returned error message is unclear, helping the model handle errors more intelligently.

A local-first, no-API-key MCP server that enables LLMs to search the web, fetch pages, and read documents using multiple engines and smart fallbacks.

Enables LLMs to interact with the AT Protocol ecosystem, including Bluesky, through natural language. Supports public data access without authentication and full write operations with authentication.