"Writing about MCP" matching MCP servers:

A Model Context Protocol server that provides AI agents with secure access to local filesystem operations, enabling reading, writing, and managing files through a standardized interface.

A persistent memory implementation using a local knowledge graph that lets Claude remember information about users across conversations.

The AgentOps MCP server provides access to observability and tracing data for debugging complex AI agent runs. This adds crucial context about where the AI agent succeeds or fails.

Enables Codex to delegate frontend development, general implementation, and writing tasks to Gemini agents through task orchestration and dedicated writing tools, with opt-in file write and native tool permissions. Tasks are supervised across MCP disconnections, allowing progress polling, follow-ups, and explicit interruption.

A transparent message ferry between Claude Desktop and Claude Code CLI, enabling task delegation with any MCP configuration without the bridge needing to know about specific MCPs.

A local, no-code builder for MCP servers that lets you create AI tools by describing them in plain language, without writing code. Connect your tools to any MCP-compatible assistant like Claude, Cursor, or VS Code Copilot.

Enables Cursor agents to communicate via a shared chat room, allowing them to ask questions, share status, and warn about conflicts while collaborating on the same repo.

Lets MCP clients invoke seven specialized agents for orchestration, coding, finance, research, design, writing, and security tasks, including multi-step tool-using jobs, and browse or read a curated 48-entry NVIDIA skills catalog.

MCP server that bridges Claude Code to Google Antigravity's CLI, enabling Claude to orchestrate Gemini workers in parallel for reading, writing, verifying, and autonomous tasks. It provides tools for single dispatch, parallel fan-out, background jobs, and handles permission, cwd, and shell pitfalls.

MCP server that enables Claude Code to communicate with other Claude Code agents over HTTP, allowing users to ask questions about remote codebases or delegate coding tasks.

CodeMesh enables AI agents to orchestrate any MCP server by writing TypeScript code, with self-improving capabilities through auto-augmentation of tool documentation.

Enables different AI assistants to exchange messages with each other across platforms and sessions using natural language commands, with persistent delivery so conversations survive restarts. Supports registering instances, sending messages to named peers, and checking inboxes without writing code.

An authless Cloudflare Workers server that provides tools to scrape, analyze, and answer questions about websites using Cloudflare's Browser Rendering and AI capabilities.

Enables interaction with S2 StreamStore to manage basins, access tokens, and data streams through a progressive tool disclosure workflow. It allows AI agents to efficiently navigate between account management, basin configuration, and stream-level operations like reading or writing records.

An MCP (Multi-Agent Conversation Protocol) Server that provides an interface to access Pokémon data through the PokéAPI, enabling agents to retrieve information about Pokémon species, abilities, moves, and more via natural language.

Enables AI assistants running on different platforms and sessions to exchange persistent messages with each other through simple natural language commands. Instances can register a name, send messages, check inboxes, and list online peers without writing any code.

An MCP server that keeps raw tool output out of the context window by sandboxing and compressing results (about 98% reduction), persisting file edits, git operations, tasks and decisions in SQLite with BM25 retrieval so the agent resumes after compaction, and enforcing routing across 17 platforms via MCP and hooks. It also encourages the model to write scripts that compute answers and log only results instead of reading large data into context.

Enables agents to accomplish multi-step work by writing a single sandboxed JavaScript program that drives shell, filesystem, HTTP, and every tool of any configured MCP server, returning only a small summary instead of raw JSON. This collapses dozens of advertised tool schemas into one tool, cutting per-request context by roughly 26× while keeping disabled upstream servers reachable.

DevTrack is a local developer-context MCP server that gives AI coding assistants access to your active ticket, commits, pending actions, writing voice profile, ticket history, and end-of-day summary. It exposes six read-only, non-destructive SQLite-backed tools over stdio. DevTrack runs locally and does not require the optional Python backend or a hosted MCP endpoint. https://devtrack.cloud.

Enables AI coding agents to get one independent reviewer that first reasons about a problem without seeing the proposed solution, then compares it against the revealed proposal and returns a compact verdict (KEEP/MODIFY/REPLACE/INSUFFICIENT_EVIDENCE) with risks, an alternative, and a falsification probe before committing to expensive decisions.

Just to be clear about the final JSON: the MCP server is named herdr-inbox, and it enables Claude Code agents in herdr to ask questions asynchronously without stalling, with all open questions answered from one popup and answers routed back to the asking agent.

Enables converting and migrating coding-agent instruction files and MCP server configurations between Claude Code, Codex CLI, Cursor and Gemini CLI, with read-only detection of existing configs, dry-run project migration, and a doc-linked compatibility matrix. Runs locally over stdio with no network calls or telemetry, writing files only when explicitly requested.

Provides AI assistants with real-time access to the AGNO framework documentation by enabling them to browse, search, and fetch documentation pages. This server allows users to query information about AGNO's Agents, Teams, and Workflows directly through MCP-compatible clients.

An unofficial MCP server that lets MCP clients like Codex operate the ChatGPT web interface through a persistent browser, enabling prompt writing, file uploads, message sending, and reading complete responses without using the ChatGPT API.

Enables Claude to delegate entire browser tasks to TypeSafe's Jev, which performs each click/type/select decision in about 100 ms, allowing fast automated web interactions while Claude handles planning and verification.

Enables Claude Code and Gemini CLI to see your macOS screen, answering questions about the file, error, UI, database table, or web page you're looking at. Captures screenshots, reads database schemas, and controls Chrome/Brave, presenting answers as on-screen highlights or spoken bubbles above your cursor.

Provides comprehensive documentation about AWS AgentCore framework to GenAI tools, enabling users to build production-ready AI agents with enterprise-grade security, observability, and scalability. Offers guidance on identity management, API integration, monitoring, code execution, memory storage, and tool integration for AI agents.

Drive Clearcote, a Chromium build with its fingerprint protection patched into the engine, from any MCP client: about 20 tools to navigate, read, extract, click, fill, screenshot and keep sessions, plus a CDP endpoint that hands the same browser to Playwright or Puppeteer. The server lives in the repo's mcp/ folder (npx clearcote-mcp or pip install clearcote-mcp).

Enables any MCP-capable agent to optimize prompts across code, writing, image, and chat scenes, returning a rewritten prompt plus change explanations and element coverage, or to diagnose a prompt with a 100-point score and issue list without rewriting it. Runs over stdio against any OpenAI-compatible endpoint configured once through environment variables or a shared config file.

Keeps large tool outputs out of an agent's context window by spilling them to disk and handing back a short pointer instead. Lets the agent pull only what it needs — exact numbered line ranges read directly from disk, or verbatim relevant lines returned by a small model answering a question about the stored output, plus tools to open, list, and close documents.

Enables an LLM to drive a real browser by declaring one outcome at a time, while Jev chooses the element, action, and value and Playwright executes it, returning a clear status such as done, blocked, or needs confirmation. Also exposes tools to open pages, ask yes/no or multiple-choice questions about a page, read question-relevant page text, capture snapshots and screenshots, act on elements directly, and invoke a page's own WebMCP tools.

Enables Claude Code to delegate subagent model routing and to answer yes/no, choice, and score questions about entire files using a small, cheap external model, so large numbers of files can be screened without filling the main context window.

Turns any REST/OpenAPI, SOAP, GraphQL, OData, SQL, MongoDB or existing MCP system into MCP tools for Claude, ChatGPT and Copilot without writing an MCP server, shipping 474 ready-made connectors with runtime tool registration, field filtering and role-based access control.

Enables Claude to offload noisy commands, large files, and web research to a local OpenAI-compatible LLM, receiving only concise cited answers instead of the raw material. It also optionally delegates writing a single file inside an isolated git worktree, retrying until a Claude-authored test passes.

Exposes ten tools that let agents and humans coordinate through a deterministic control plane: starting tasks, acquiring and releasing typed file leases, writing and searching permissioned memory, requesting approvals, polling structured events, and submitting work through a merge gate. Every action is checked against signed mandates and policy before it succeeds, so multi-agent teams can share a repo and an audit trail without a supervisor model watching every session.

Exposes a single execute tool so agents can discover, inspect schemas for, and compose calls to connected upstream MCP servers by writing JavaScript that runs in a fresh sandboxed QuickJS VM. Results from combined tool calls are returned as text or JSON, keeping large tool catalogs out of the agent's context.

Enables MCP clients to classify plain-language messages against runtime-defined intents, with margin-based doubt flags, beta yes/no checks, and per-intent probability scores, without training examples. It runs self-hosted and supports multilingual intent decisions in about 26 ms per message on a compiled GPU.

Enables agents to report a failed command so it is matched against a shared Postgres registry of crash sites and answered with a briefing of other agents' confirmed fixes, before or during a task. Also exposes proven routes for upcoming work, vendor airworthiness ratings, and dependency-based preflight warnings, so failures are warned about rather than rediscovered.

Enables multiple AI agents and IDEs to coordinate through a shared Markdown vault by delegating, claiming, and completing tasks and by writing and searching shared memory. It keeps the task dashboard synchronized while allowing safe cross-process access to the same workspace.

Enables AI agents to browse, click, type, screenshot, and extract data from web pages without writing CSS selectors, using numbered element refs. It also supports office automation such as bulk form filling, table extraction, downloads, PDF archiving, and page monitoring.