"Loop" matching MCP servers:

Provides bounded MCP tools for AI agents to inspect, write, verify, and diff files in a sandboxed repository, requiring human approval and logging an audit trail.

Enables secure file system operations (read, write, delete) and simulated command execution with server-enforced permission policies, risk assessment, and human-in-the-loop approval.

Fill any PDF form with AI agents — ML field detection on scans, visual review loop, reusable templates, and native AcroForm fill via justfill.app.

Closes the find-download-install loop for lutris-mcp by searching Prowlarr, downloading via qBittorrent, polling until complete, classifying the result, and returning a path for installation.

MCP server for automated visual-layout QA of DOCX documents, enabling rendering, measurement, rule checking, patching, and diffing in a closed loop.

Provides comprehensive computer control capabilities including shell commands, file operations, screen capture, keyboard/mouse input, window management, clipboard access, and system monitoring following the OODA loop pattern.

A custom MCP server for orchestrating Windows/Linux/macOS workstations, providing filesystem, shell, SSH, and personal productivity tools with security features like allowlisting and human-in-the-loop approval for writes.

Governed execution cells for AI agents via MCP; provides shell, code, file, and browser tools inside isolated, policy-enforced Citadels with deny-by-default security, human-in-the-loop approvals, and tamper-evident audit trails.

Enables authenticated web agents to reach a local machine through a portable MCP endpoint, with optional Cloudflare Tunnel or Tailscale Funnel ingress for public access.

An autonomous DFIR agent that hunts timestomping and rename/move laundering on the SANS SIFT Workstation, driven by Claude Code, with a read-only architecture and self-correcting triage loop using the USN change journal.

Provides a human-in-the-loop security layer for AI agents by intercepting file operations, explaining them with a local LLM, and enforcing a deterministic policy that requires user approval for risky actions.

Read-only MCP server for autonomous, spoliation-proof disk-image triage with self-correcting verification loop.

Enables conversational processing of Daily Status Reports (DSR) by extracting, validating, and storing data from JSON, DOCX, and XLSX files in IBM Cloud Object Storage. It provides a human-in-the-loop workflow for batch file processing and schema validation directly within MCP-compatible clients like Claude Desktop.

MCP server for reading and listing JSON documents, designed to be used with Claude Haiku and Headroom compression to reduce token usage. It demonstrates an end-to-end MCP tool-use loop with stdio transport.

MCP server that provides secure read-only access to a local folder, enabling file listing, reading, semantic search (RAG), and indexing status via natural language, integrated with Claude Desktop and a custom agent loop.

Web-based SSH connection manager with an MCP server that enables AI agents to manage remote hosts via terminals, file transfers, and port forwarding, featuring human-in-the-loop approval for secure access.

MCP server for AI-powered resume optimization and generation. It enables clients to optimize existing resumes or generate new ones from background info via LLM, producing formatted .docx and .pdf files.

MCP server for Word document (.docx) creation and manipulation — the production-grade document automation tool for AI agents.

Provides full control of a macOS system, enabling AI agents to execute shell commands, run AppleScript, manage files and processes, and perform human-like GUI interactions such as clicking, dragging, scrolling, and typing, with a screenshot-based perception loop.

Enables safe, human-in-the-loop AI document collaboration by letting agents create isolated workspaces, edit tracked files, open reviewable Change Requests, and get human approval before merging, with Git-backed conflict detection and resolution.