"A desktop commander that uses UVX" matching MCP connectors:

Drive real Android & iOS devices and web browsers from natural language for mobile + web QA. 145+ tools across device control, app management, automation sessions, browser automation, and flow recording / replay. Bearer-auth — get a token at robotactions.com → Profile → API Tokens.

Prompt evals over MCP: run a prompt on your dataset, score each output 1-5 with an LLM judge.

Check that your AI is being logical. Free tool that mathematically catches contradictions in agent reasoning. No account needed. Also offers paid guardrails that converts natural language to formal verification proofs, that anyone can check succinctly.

A skeptical senior-engineer code reviewer over MCP: risk-scans unified diffs, flags AI-generated-code tells, reports complexity hotspots, scans for leaked secrets, and runs an OWASP security pass — real analyzers, no external APIs. Free tier, no signup.

A fully free linter for agent skill files: lint_skill validates YAML frontmatter, structure, size budgets, and safety phrasing with a pass/fail verdict; packaging_check validates zip layout against marketplace rules; plus regex_test, json_validate, diff_texts, and cron_explain for skill authors. No license or account required.

Score any URL against a real design contract — 40 checks, A-F grade, token + motion validation.

Scan any website or MCP server for agent readiness: 0-100 score, a fix per failing check. Free.

Post-scrape data cleaner, no LLM: repairs mojibake, HTML, invisible chars. Plus a verdict.

Benchmark-first release surface with a read-only MCP endpoint and operator CLI.

A webhook inbox for agents: one call returns a live URL. Mock, verify, inspect and replay.

Accessibility pre-checks (WCAG/BFSG) in a real browser + statement drafts. Pay per call.

Grade MCP servers A to F with the open behavioral litmus. npm: full toolset; hosted: lookups only.

Hire a real human for real-world verification, product testing, AI output review, and errands.

Runs your code against a contract; returns HELD or BROKE at the exact input. Deterministic.

MCP server for the Fail Modes taxonomy — a knowledge base of AI system failure modes

A flock of AI users tests your deployed app and reports where real people get stuck, with fixes.

Real-browser WCAG audit that also finds keyboard-inoperable controls axe-core misses, with fixes.

SeaOtter dispatches work to a Superteam and checks the delivered outcome before money moves.

AI QA that runs your app in a browser on every pull request: projects, test targets, test cases.

Disposable test mailboxes on a real domain: send, receive and assert on real email.