Airsc is free for everyone. If it helps you, support it with $2–$20.

Support
VISign in

Best Testing & debugging resources for vibe coders

Updated 09/10/2026

This guide is for people who build with AI agents and do not want to read every line of code. The agent says it works. These resources help you check that it really does.

They cover three jobs: proving the agent's tests are real, running your app in a browser to look for bugs, and tracing errors back to the code. Start with the beginner picks first. Some tools are marked caution, so read what they access before you install.

  1. PluginSafe

    dotpals

    Rikinshah787

    This plugin helps users see what their coding agent actually did, verifies test passes, and makes agents fix failing tests.

    36MIT

    #1 · Why we picked it

    Beginner level, safe label and the top quality score (98); it catches an agent faking a test pass and pushes it to fix failing tests.
  2. PluginSafe

    verify

    opslane

    This Claude Code plugin helps you verify agent-generated code by booting your application, driving changes against approved acceptance criteria, and providing a detailed report.

    115MIT

    #2 · Why we picked it

    Safe and well rated (92); it boots your app and checks the agent's changes against acceptance criteria you approve, then gives a report.
  3. PluginSafe

    playwright-skill

    lackeyjb

    A skill that enables coding agents to write and execute custom Playwright scripts for complex web flows.

    3.2kMIT

    #3 · Why we picked it

    Safe label; lets the agent write and run Playwright scripts so you can watch the browser session and record video or network logs.
  4. SkillSafe

    SeleniumBase

    seleniumbase

    An MCP server and framework for browser automation, web scraping, and end-to-end testing with stealth capabilities.

    13kMIT

    #4 · Why we picked it

    Safe, 13,000+ stars, and covers UI tests and scraping through an MCP server; expect some setup and intermediate-level work.
  5. MCP serverSafe

    httprunner

    httprunner

    HttpRunner is an open-source API/UI testing framework that uses LLM technology for natural language driven test scenarios, supporting multiple platforms and offering broad test capabilities.

    4.3kApache-2.0

    #5 · Why we picked it

    Safe; turns plain-language descriptions into API test cases and can reuse them for load testing.
  6. PluginCaution

    toolkit

    getsentry

    Provides agentic tooling as an MCP server and Claude plugin to integrate Sentry's error monitoring and performance debugging features into coding agents.

    913

    #6 · Why we picked it

    Quality 88; lets Claude Code read Sentry errors and suggest fixes, but it is marked caution and you need a Sentry project to use it.
  7. MCP serverCaution

    vibetest-use

    browser-use

    This MCP server launches browser agents to perform automated QA testing on websites for UI bugs, broken links, and accessibility issues.

    833

    #7 · Why we picked it

    Launches browser agents to hunt for UI bugs, broken links and accessibility issues on a local or live site; marked caution, so check its setup and permissions.
  8. MCP serverCaution

    inspector

    MCPJam

    MCPJam is a testing and evaluation platform to chat with, inspect, and debug MCP servers and AI agent applications across various client configurations.

    2.2k

    #8 · Why we picked it

    Beginner level; helps you inspect and debug an MCP server's tool calls with full traces, but it is marked caution.

How to choose

  • If you only trust what you can see, start with dotpals or verify. They check the agent's own claims.
  • For web apps, pick one browser tool: playwright-skill if you want to watch it run, SeleniumBase if you also need scraping.
  • For errors in a live app, use getsentry-toolkit. It needs a Sentry account and has a caution label.
  • Prefer the safe label when you have a choice. For caution tools, read the permissions and try them on a test project first.
  • Skip advanced reverse-engineering tools like x64dbg or Cheat Engine bridges unless you debug binaries on purpose.

FAQ

How do I know if my AI agent faked a test pass?

dotpals is built for this. It shows what the agent actually did, which files it changed and which commands it ran. verify also boots your app and checks the result against your criteria.

Which tool should a complete beginner try first?

dotpals. It is rated beginner, has a safe label and the highest quality score in this list. Add playwright-skill later when you want browser tests.

Can these tools test my app without me writing test code?

Mostly yes. httprunner builds API tests from plain-language descriptions, and vibetest-use sends browser agents to look for UI bugs and broken links. You still need to read the reports.

What does the caution label mean for me?

It means the resource needs a closer look before use. Read what it can access, avoid giving it real secrets, and try it on a throwaway project first.