Development MCP Servers — Page 3

Read-only MCP and x402 endpoint preflight checks: 402 challenge, payment documents, accepts[] conformance, discovery manifests, response digest. Zero dependencies.

Security-first local SSH and database MCP server with pre-dispatch safety guardrails, credential isolation, and atomic file deployments.

Let AI agents act for a person on websites, with the person's approval on their phone. Take over lets the owner drive a stuck agent's browser from their phone.

A real cloud Mac with Xcode for your agent: start macOS, run commands, build and test iOS apps, ship to TestFlight.

Run a GPU job under a hard dollar cap and collect the files it writes.

Live B2B GTM data API: people, companies, emails, phones, web intelligence. One key. On waterfall endpoints a miss costs zero credits.

Screenshot Diff by lintlab: capture full-page, viewport or element screenshots of public URLs and optionally compare them against a baseline page or image, returning a diff image and the percentage of pixels changed. A pay-per-event Apify Actor, used as an MCP tool through Apify's hosted MCP server; needs an Apify token.

Free MCP tools for prompt-injection test inputs, OWASP LLM risk mapping, AI release planning, and custom response test-pack creation for PromptBrake CI.

Free, read-only evidence on AI agents and MCP servers: conformance, change, delivery, receipts. No account or API key.

An open-source benchmark of how well LLMs solve nonogram puzzles, from 5x5 to 20x20.

LinkedIn, Sales Navigator, Recruiter, WhatsApp, Instagram, Telegram and email MCP for coding agents.

Link infrastructure for mobile teams: deep links, install attribution, and real-time analytics. Create and update links, register apps, pull analytics, and manage keys and custom domains from any MCP client. Remote server: https://api.warplink.app/mcp (OAuth sign-in or API key).

TaskForceAI helps small teams manage AI agents, track tasks and review results, bringing AI into everyday workflows with human oversight.

Helps coding agents pick dependencies with evidence. Give it a need in plain words ("python pdf parser") and it returns GitHub repositories ranked by a transparent 0-100 score (momentum, maintenance, npm/PyPI downloads, license) with a one-line verdict each. Tools: recommend_repos, compare_repos, trending_repos. Install: npx -y whichlib

Live web checks for AI agents: sitemaps, robots.txt, URL status, broken links, feeds, citations.

A personal lexicon for voice-to-agents. One YAML file of the words speech-to-text gets wrong, applied everywhere your voice lands: MCP, Claude Code, the browser, macOS.

Free, read-only email and DNS checks: SPF, DKIM, DMARC, MX, DNS lookup, blacklists and a full email domain health check, each with a link to the full report on mxverdict.com.

Voice endpoint heuristics using supplied frame energies, transcript fragments and silence duration. Local Python stdio MCP; does not process raw audio.

Apply transcript, supplied echo score and duration rules to conversational interruptions and estimate playback truncation. No audio echo detection or playback control.

Local Python MCP tools for packet metadata ordering and jitter buffer simulation. No audio codec or WebRTC network transport.

Record supplied voice pipeline timestamps and calculate latency, bottlenecks and nearest-rank percentiles. Local in-memory MCP; no automatic instrumentation.

Calibrate linear chart axes and convert supplied bar or scatter pixel coordinates to values. Local Python MCP; no image recognition or logarithmic axes.

Deterministic grounding checks for RAG and document-QA answers: required facts present, every number traceable to the context, and abstention when the context cannot answer. No LLM judge.

Scaffolds a deployment-readiness harness (redaction, evals from real failures, CI regression gate) into document-QA and extraction projects, and exposes grounding, extraction and readiness checks to agents.

Audit a webpage's AI crawler access from an MCP-compatible client. The audit_url tool returns a score and evidence-based findings; get_quota returns the available audit capacity. Remote endpoint: https://app.nyman.media/api/mcp. Transport: Streamable HTTP. Authentication: bearer API key. MCP access: Premium or Partner subscription.

Connect your coding agent to QA Wolf to create, run, and troubleshoot end-to-end tests. Explore your app in a cloud browser, review a proposed test outline, and approve QA Wolf to implement and validate it. Sign in with your QA Wolf account and follow progress through live session links.

Website monitoring for any site, deepest on WordPress & WooCommerce: add sites, run checks, read results and get fix prompts.

Connects the AI agents you use: your agent hands a task to another and gets the answer back

Upload an .apk or .ipa and get an install link and QR code your testers open on their phone. A folder for every branch, iOS ad hoc installs with device registration, and tester feedback with screenshots. Works with Expo, React Native, Flutter and native apps, from the dashboard, a CLI, CI or an AI coding agent. Free tier; paid plans from $9 a month.

Hubris is an OpenAI-compatible LLM gateway billed in rubles: one API key for 500+ models from OpenAI, Anthropic, Google, DeepSeek, Qwen and others. The remote MCP server lets agents search the model catalogue, compare prices, check the account balance and run chat completions on any model.