Development MCP Servers

Your coding agent finishes a feature and hands you a UAT link. You walk it on the real app and pin what is wrong. The agent reads each pin, fixes it and reports back. Only a person marks a fix verified. 6 tools, OAuth sign-in, works with Claude Code and any MCP client.

Free, self-hosted WordPress MCP server with per-change undo. Claude, ChatGPT, Gemini, Grok, Perplexity, Cursor, Claude Code or any MCP client manage posts, pages, WooCommerce, plugins, users, media, menus, files, the database and WP-CLI on a live site, with safe mode, dry run, database checkpoints and an audit log. Up to 203 tools behind two gateway tools, one-click browser sign-in (OAuth 2.1), no relay, no account, no Pro tier. Installed from WordPress.org.

Generate PDFs and PNG images from HTML, Markdown, URLs or Handlebars templates; free API key with 100 renders a month.

Turn a client mockup into a native, editable Elementor V4 page from Claude, ChatGPT or Cursor: design tokens, SEO audit, Schema.org. Paid WordPress plugin.

Zero-telemetry local developer execution layer for AI coding agents. Provides deterministic tools to inspect listening network ports and processes, calculate IPv4 CIDR subnet ranges and usable hosts, decode embedded creation timestamps from distributed IDs (UUIDv7, ULID, Snowflake, MongoDB ObjectId), and convert Unix epoch timestamps to UTC, ISO 8601, and local time.

Local privacy testing for AI agents. Scan code and captured traces across tool calls, memory, logs and output; redact sensitive data.

Continuous mandate enforcement for AI agents: every proposed action is checked against a signed mandate BEFORE it happens. Six rules in fixed order (amount, payee, audience, currency, window, intent), deny-by-default, tamper-evident ledger fingerprint. HTTP + MCP endpoint, free tier 100 checks/day. Apache-2.0.

Instant website audits: SEO, speed, security headers, SSL, mobile and broken links. Free web scan; pay-per-call MCP tools for agents.

Search and read 420+ web development, JavaScript, CSS, canvas, animation, and DSA tutorials from KIRUPA.com.

Verify agent artifacts before you trust them: schema, emptiness, prompt-injection and provenance checks in four layers, deny-by-default. HTTP + MCP endpoint, free tier 20 checks/day, paid per-call via x402 on Base.

MCP proxy that records every AI tool call in a tamper-evident SHA-256 hash chain, with deny-by-default allowlist, optional human approval for writes and fail-closed behaviour. Designed to support GLP / 21 CFR Part 11 / EU GMP Annex 11 workflows. MIT.

Lets coding agents read CI failures self-heal couldn't fix, check run status, dispatch workflows and run jobs on fresh Linux VMs.

Hand Rush a task from any MCP client: an agent does the work in its own cloud sandbox and returns a PR, report or files, with the full trajectory.

Public URLs for AI agents to catch webhooks, OAuth redirects, job callbacks and approval clicks.

Capture web pages as PNG, JPEG or WebP, render a URL or HTML to PDF, and generate 1200x630 OG images. REST API and remote MCP server. Free tier.

One MCP endpoint to a live API catalog. AI agents pay per call in USDC (x402 on Base, MPP on Tempo).

Read Layerdock website feedback in your coding agent. Each pin comes with the CSS selector, screenshot, page URL and the console and network trace captured when the client reported it. Fix the bug, reply on the thread and mark it resolved without leaving the editor. Works with Claude Code, Cursor, Codex CLI, Windsurf, VS Code, Zed and Antigravity.

Find the cheapest AI model that passes your own tests: audit the models your repo uses, draft evals, rank by cost per passing answer, and generate cheapest-first routing. Local, no API keys.

Check whether an IP is residential, datacenter or proxy, and whether ChatGPT, Claude, Gemini or Muse support its region.

Audit a domain's email deliverability: SPF, DKIM, DMARC, MX, mail provider, catch-all, blacklist, domain age, and a 0 to 100 health score. Clay-ready output.

Local AI software studio with mission, approval, and self-healing QA tools through a Python stdio MCP server.

Check x402 and MPP paid APIs before your agent pays: free endpoint check, $0.001 go/caution/no_go preflight, and fixes for sellers.

Free, read-only remote MCP server (no auth) over the AgentTask Atlas catalog: MCP servers and Agent Skills organized by dev task, each with a one-line practical verdict, install command, credentials and source link. Tools: search_catalog, get_entry, list_categories.

Search Apple's official developer documentation and WWDC transcripts with hybrid RAG retrieval and Jev relevance ranking.

Review videos from your coding agent: pick a style, then point at the frame and say what to change. Works with Hyperframes, Remotion and any other video tool, over MCP.

Omnara is an open-source managed agent platform. Its MCP server lets Claude Code, Codex, Cursor, VS Code, and other MCP clients launch Omnara agents on any model and machine, message them, read their events, and approve their tool calls, with the permissions of your Omnara account. Sign in through the browser with OAuth; there's nothing to install.

Dated, verified fixes for how APIs, platforms and AI coding tools break, served to your coding agent before it writes the code.

Five small tools for agent builders: AI-writing tells, tools/list token cost, CLAUDE.md lint, install snippets, AI crawler check.

One room for Claude Code, Codex, Gemini CLI and opencode agents: @-mention one to fix something in its own git worktree, another reviews it, and it publishes to your repo. Runs locally; MIT.

MCP server that lets an AI assistant list, health-check, connect and install APKs on many Android devices over ADB in parallel (USB, emulators, ADB-over-TCP, cloud phones). Shell access is off by default.